
import numpy as np
import matplotlib
import matplotlib.pyplot as plt

import plotly.graph_objects as go
import plotly.express as px
from plotly.subplots import make_subplots

import pandas as pd
import dtale
import sweetviz as sv
import pandas_profiling as pdp
from sklearn import datasets
from sklearn.svm import SVC

from sklearn.svm import SVC
from sklearn.gaussian_process.kernels import RBF
from sklearn.linear_model import LogisticRegression
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import RandomForestClassifier
from sklearn.ensemble import GradientBoostingClassifier
from sklearn.ensemble import VotingClassifier
from xgboost import XGBClassifier
# from lightgbm import LGBMClassifier
from catboost import CatBoostClassifier

from sklearn.model_selection import train_test_split
from sklearn.preprocessing import MinMaxScaler
from sklearn.metrics import classification_report
from sklearn.metrics import confusion_matrix
from sklearn.metrics import accuracy_score
from sklearn.metrics import precision_score
from sklearn.metrics import recall_score
from sklearn.metrics import f1_score
from sklearn.metrics import roc_auc_score
from sklearn.metrics import roc_curve

import missingno as msno
import seaborn as sns
%matplotlib inline
 
pd.options.display.max_columns = None
pd.options.display.max_rows = None


raw_data = pd.read_csv('covid_data.csv', header = 0)


raw_data.head()


raw_data.tail()


raw_data.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 1925 entries, 0 to 1924
Columns: 231 entries, PATIENT_VISIT_IDENTIFIER to ICU
dtypes: float64(225), int64(4), object(2)
memory usage: 3.4+ MB


number_of_patients = raw_data['PATIENT_VISIT_IDENTIFIER'].max()+1
print(f"The total number of patients from the dataset is {number_of_patients}")

The total number of patients from the dataset is 385


msno.bar(raw_data)

<AxesSubplot:>


print('NaN values =', raw_data.isnull().sum().sum())
print("""""")

vars_with_missing = []

for feature in raw_data.columns:
    missings = raw_data[feature].isna().sum()
    
    if missings > 0 :
        vars_with_missing.append(feature)
        missings_perc = missings / raw_data.shape[0]
        
        print('Variable {} has {} records ({:.2%}) with missing values.'.format(feature, missings, missings_perc))
print('In total, there are {} variables with missing values'.format(len(vars_with_missing)))

NaN values = 223863

Variable DISEASE GROUPING 1 has 5 records (0.26%) with missing values.
Variable DISEASE GROUPING 2 has 5 records (0.26%) with missing values.
Variable DISEASE GROUPING 3 has 5 records (0.26%) with missing values.
Variable DISEASE GROUPING 4 has 5 records (0.26%) with missing values.
Variable DISEASE GROUPING 5 has 5 records (0.26%) with missing values.
Variable DISEASE GROUPING 6 has 5 records (0.26%) with missing values.
Variable HTN has 5 records (0.26%) with missing values.
Variable IMMUNOCOMPROMISED has 5 records (0.26%) with missing values.
Variable OTHER has 5 records (0.26%) with missing values.
Variable ALBUMIN_MEDIAN has 1104 records (57.35%) with missing values.
Variable ALBUMIN_MEAN has 1104 records (57.35%) with missing values.
Variable ALBUMIN_MIN has 1104 records (57.35%) with missing values.
Variable ALBUMIN_MAX has 1104 records (57.35%) with missing values.
Variable ALBUMIN_DIFF has 1104 records (57.35%) with missing values.
Variable BE_ARTERIAL_MEDIAN has 1104 records (57.35%) with missing values.
Variable BE_ARTERIAL_MEAN has 1104 records (57.35%) with missing values.
Variable BE_ARTERIAL_MIN has 1104 records (57.35%) with missing values.
Variable BE_ARTERIAL_MAX has 1104 records (57.35%) with missing values.
Variable BE_ARTERIAL_DIFF has 1104 records (57.35%) with missing values.
Variable BE_VENOUS_MEDIAN has 1104 records (57.35%) with missing values.
Variable BE_VENOUS_MEAN has 1104 records (57.35%) with missing values.
Variable BE_VENOUS_MIN has 1104 records (57.35%) with missing values.
Variable BE_VENOUS_MAX has 1104 records (57.35%) with missing values.
Variable BE_VENOUS_DIFF has 1104 records (57.35%) with missing values.
Variable BIC_ARTERIAL_MEDIAN has 1104 records (57.35%) with missing values.
Variable BIC_ARTERIAL_MEAN has 1104 records (57.35%) with missing values.
Variable BIC_ARTERIAL_MIN has 1104 records (57.35%) with missing values.
Variable BIC_ARTERIAL_MAX has 1104 records (57.35%) with missing values.
Variable BIC_ARTERIAL_DIFF has 1104 records (57.35%) with missing values.
Variable BIC_VENOUS_MEDIAN has 1104 records (57.35%) with missing values.
Variable BIC_VENOUS_MEAN has 1104 records (57.35%) with missing values.
Variable BIC_VENOUS_MIN has 1104 records (57.35%) with missing values.
Variable BIC_VENOUS_MAX has 1104 records (57.35%) with missing values.
Variable BIC_VENOUS_DIFF has 1104 records (57.35%) with missing values.
Variable BILLIRUBIN_MEDIAN has 1104 records (57.35%) with missing values.
Variable BILLIRUBIN_MEAN has 1104 records (57.35%) with missing values.
Variable BILLIRUBIN_MIN has 1104 records (57.35%) with missing values.
Variable BILLIRUBIN_MAX has 1104 records (57.35%) with missing values.
Variable BILLIRUBIN_DIFF has 1104 records (57.35%) with missing values.
Variable BLAST_MEDIAN has 1104 records (57.35%) with missing values.
Variable BLAST_MEAN has 1104 records (57.35%) with missing values.
Variable BLAST_MIN has 1104 records (57.35%) with missing values.
Variable BLAST_MAX has 1104 records (57.35%) with missing values.
Variable BLAST_DIFF has 1104 records (57.35%) with missing values.
Variable CALCIUM_MEDIAN has 1104 records (57.35%) with missing values.
Variable CALCIUM_MEAN has 1104 records (57.35%) with missing values.
Variable CALCIUM_MIN has 1104 records (57.35%) with missing values.
Variable CALCIUM_MAX has 1104 records (57.35%) with missing values.
Variable CALCIUM_DIFF has 1104 records (57.35%) with missing values.
Variable CREATININ_MEDIAN has 1104 records (57.35%) with missing values.
Variable CREATININ_MEAN has 1104 records (57.35%) with missing values.
Variable CREATININ_MIN has 1104 records (57.35%) with missing values.
Variable CREATININ_MAX has 1104 records (57.35%) with missing values.
Variable CREATININ_DIFF has 1104 records (57.35%) with missing values.
Variable FFA_MEDIAN has 1104 records (57.35%) with missing values.
Variable FFA_MEAN has 1104 records (57.35%) with missing values.
Variable FFA_MIN has 1104 records (57.35%) with missing values.
Variable FFA_MAX has 1104 records (57.35%) with missing values.
Variable FFA_DIFF has 1104 records (57.35%) with missing values.
Variable GGT_MEDIAN has 1104 records (57.35%) with missing values.
Variable GGT_MEAN has 1104 records (57.35%) with missing values.
Variable GGT_MIN has 1104 records (57.35%) with missing values.
Variable GGT_MAX has 1104 records (57.35%) with missing values.
Variable GGT_DIFF has 1104 records (57.35%) with missing values.
Variable GLUCOSE_MEDIAN has 1104 records (57.35%) with missing values.
Variable GLUCOSE_MEAN has 1104 records (57.35%) with missing values.
Variable GLUCOSE_MIN has 1104 records (57.35%) with missing values.
Variable GLUCOSE_MAX has 1104 records (57.35%) with missing values.
Variable GLUCOSE_DIFF has 1104 records (57.35%) with missing values.
Variable HEMATOCRITE_MEDIAN has 1104 records (57.35%) with missing values.
Variable HEMATOCRITE_MEAN has 1104 records (57.35%) with missing values.
Variable HEMATOCRITE_MIN has 1104 records (57.35%) with missing values.
Variable HEMATOCRITE_MAX has 1104 records (57.35%) with missing values.
Variable HEMATOCRITE_DIFF has 1104 records (57.35%) with missing values.
Variable HEMOGLOBIN_MEDIAN has 1104 records (57.35%) with missing values.
Variable HEMOGLOBIN_MEAN has 1104 records (57.35%) with missing values.
Variable HEMOGLOBIN_MIN has 1104 records (57.35%) with missing values.
Variable HEMOGLOBIN_MAX has 1104 records (57.35%) with missing values.
Variable HEMOGLOBIN_DIFF has 1104 records (57.35%) with missing values.
Variable INR_MEDIAN has 1104 records (57.35%) with missing values.
Variable INR_MEAN has 1104 records (57.35%) with missing values.
Variable INR_MIN has 1104 records (57.35%) with missing values.
Variable INR_MAX has 1104 records (57.35%) with missing values.
Variable INR_DIFF has 1104 records (57.35%) with missing values.
Variable LACTATE_MEDIAN has 1104 records (57.35%) with missing values.
Variable LACTATE_MEAN has 1104 records (57.35%) with missing values.
Variable LACTATE_MIN has 1104 records (57.35%) with missing values.
Variable LACTATE_MAX has 1104 records (57.35%) with missing values.
Variable LACTATE_DIFF has 1104 records (57.35%) with missing values.
Variable LEUKOCYTES_MEDIAN has 1104 records (57.35%) with missing values.
Variable LEUKOCYTES_MEAN has 1104 records (57.35%) with missing values.
Variable LEUKOCYTES_MIN has 1104 records (57.35%) with missing values.
Variable LEUKOCYTES_MAX has 1104 records (57.35%) with missing values.
Variable LEUKOCYTES_DIFF has 1104 records (57.35%) with missing values.
Variable LINFOCITOS_MEDIAN has 1104 records (57.35%) with missing values.
Variable LINFOCITOS_MEAN has 1104 records (57.35%) with missing values.
Variable LINFOCITOS_MIN has 1104 records (57.35%) with missing values.
Variable LINFOCITOS_MAX has 1104 records (57.35%) with missing values.
Variable LINFOCITOS_DIFF has 1104 records (57.35%) with missing values.
Variable NEUTROPHILES_MEDIAN has 1104 records (57.35%) with missing values.
Variable NEUTROPHILES_MEAN has 1104 records (57.35%) with missing values.
Variable NEUTROPHILES_MIN has 1104 records (57.35%) with missing values.
Variable NEUTROPHILES_MAX has 1104 records (57.35%) with missing values.
Variable NEUTROPHILES_DIFF has 1104 records (57.35%) with missing values.
Variable P02_ARTERIAL_MEDIAN has 1104 records (57.35%) with missing values.
Variable P02_ARTERIAL_MEAN has 1104 records (57.35%) with missing values.
Variable P02_ARTERIAL_MIN has 1104 records (57.35%) with missing values.
Variable P02_ARTERIAL_MAX has 1104 records (57.35%) with missing values.
Variable P02_ARTERIAL_DIFF has 1104 records (57.35%) with missing values.
Variable P02_VENOUS_MEDIAN has 1104 records (57.35%) with missing values.
Variable P02_VENOUS_MEAN has 1104 records (57.35%) with missing values.
Variable P02_VENOUS_MIN has 1104 records (57.35%) with missing values.
Variable P02_VENOUS_MAX has 1104 records (57.35%) with missing values.
Variable P02_VENOUS_DIFF has 1104 records (57.35%) with missing values.
Variable PC02_ARTERIAL_MEDIAN has 1104 records (57.35%) with missing values.
Variable PC02_ARTERIAL_MEAN has 1104 records (57.35%) with missing values.
Variable PC02_ARTERIAL_MIN has 1104 records (57.35%) with missing values.
Variable PC02_ARTERIAL_MAX has 1104 records (57.35%) with missing values.
Variable PC02_ARTERIAL_DIFF has 1104 records (57.35%) with missing values.
Variable PC02_VENOUS_MEDIAN has 1104 records (57.35%) with missing values.
Variable PC02_VENOUS_MEAN has 1104 records (57.35%) with missing values.
Variable PC02_VENOUS_MIN has 1104 records (57.35%) with missing values.
Variable PC02_VENOUS_MAX has 1104 records (57.35%) with missing values.
Variable PC02_VENOUS_DIFF has 1104 records (57.35%) with missing values.
Variable PCR_MEDIAN has 1104 records (57.35%) with missing values.
Variable PCR_MEAN has 1104 records (57.35%) with missing values.
Variable PCR_MIN has 1104 records (57.35%) with missing values.
Variable PCR_MAX has 1104 records (57.35%) with missing values.
Variable PCR_DIFF has 1104 records (57.35%) with missing values.
Variable PH_ARTERIAL_MEDIAN has 1104 records (57.35%) with missing values.
Variable PH_ARTERIAL_MEAN has 1104 records (57.35%) with missing values.
Variable PH_ARTERIAL_MIN has 1104 records (57.35%) with missing values.
Variable PH_ARTERIAL_MAX has 1104 records (57.35%) with missing values.
Variable PH_ARTERIAL_DIFF has 1104 records (57.35%) with missing values.
Variable PH_VENOUS_MEDIAN has 1104 records (57.35%) with missing values.
Variable PH_VENOUS_MEAN has 1104 records (57.35%) with missing values.
Variable PH_VENOUS_MIN has 1104 records (57.35%) with missing values.
Variable PH_VENOUS_MAX has 1104 records (57.35%) with missing values.
Variable PH_VENOUS_DIFF has 1104 records (57.35%) with missing values.
Variable PLATELETS_MEDIAN has 1104 records (57.35%) with missing values.
Variable PLATELETS_MEAN has 1104 records (57.35%) with missing values.
Variable PLATELETS_MIN has 1104 records (57.35%) with missing values.
Variable PLATELETS_MAX has 1104 records (57.35%) with missing values.
Variable PLATELETS_DIFF has 1104 records (57.35%) with missing values.
Variable POTASSIUM_MEDIAN has 1104 records (57.35%) with missing values.
Variable POTASSIUM_MEAN has 1104 records (57.35%) with missing values.
Variable POTASSIUM_MIN has 1104 records (57.35%) with missing values.
Variable POTASSIUM_MAX has 1104 records (57.35%) with missing values.
Variable POTASSIUM_DIFF has 1104 records (57.35%) with missing values.
Variable SAT02_ARTERIAL_MEDIAN has 1104 records (57.35%) with missing values.
Variable SAT02_ARTERIAL_MEAN has 1104 records (57.35%) with missing values.
Variable SAT02_ARTERIAL_MIN has 1104 records (57.35%) with missing values.
Variable SAT02_ARTERIAL_MAX has 1104 records (57.35%) with missing values.
Variable SAT02_ARTERIAL_DIFF has 1104 records (57.35%) with missing values.
Variable SAT02_VENOUS_MEDIAN has 1104 records (57.35%) with missing values.
Variable SAT02_VENOUS_MEAN has 1104 records (57.35%) with missing values.
Variable SAT02_VENOUS_MIN has 1104 records (57.35%) with missing values.
Variable SAT02_VENOUS_MAX has 1104 records (57.35%) with missing values.
Variable SAT02_VENOUS_DIFF has 1104 records (57.35%) with missing values.
Variable SODIUM_MEDIAN has 1104 records (57.35%) with missing values.
Variable SODIUM_MEAN has 1104 records (57.35%) with missing values.
Variable SODIUM_MIN has 1104 records (57.35%) with missing values.
Variable SODIUM_MAX has 1104 records (57.35%) with missing values.
Variable SODIUM_DIFF has 1104 records (57.35%) with missing values.
Variable TGO_MEDIAN has 1104 records (57.35%) with missing values.
Variable TGO_MEAN has 1104 records (57.35%) with missing values.
Variable TGO_MIN has 1104 records (57.35%) with missing values.
Variable TGO_MAX has 1104 records (57.35%) with missing values.
Variable TGO_DIFF has 1104 records (57.35%) with missing values.
Variable TGP_MEDIAN has 1104 records (57.35%) with missing values.
Variable TGP_MEAN has 1104 records (57.35%) with missing values.
Variable TGP_MIN has 1104 records (57.35%) with missing values.
Variable TGP_MAX has 1104 records (57.35%) with missing values.
Variable TGP_DIFF has 1104 records (57.35%) with missing values.
Variable TTPA_MEDIAN has 1104 records (57.35%) with missing values.
Variable TTPA_MEAN has 1104 records (57.35%) with missing values.
Variable TTPA_MIN has 1104 records (57.35%) with missing values.
Variable TTPA_MAX has 1104 records (57.35%) with missing values.
Variable TTPA_DIFF has 1104 records (57.35%) with missing values.
Variable UREA_MEDIAN has 1104 records (57.35%) with missing values.
Variable UREA_MEAN has 1104 records (57.35%) with missing values.
Variable UREA_MIN has 1104 records (57.35%) with missing values.
Variable UREA_MAX has 1104 records (57.35%) with missing values.
Variable UREA_DIFF has 1104 records (57.35%) with missing values.
Variable DIMER_MEDIAN has 1104 records (57.35%) with missing values.
Variable DIMER_MEAN has 1104 records (57.35%) with missing values.
Variable DIMER_MIN has 1104 records (57.35%) with missing values.
Variable DIMER_MAX has 1104 records (57.35%) with missing values.
Variable DIMER_DIFF has 1104 records (57.35%) with missing values.
Variable BLOODPRESSURE_DIASTOLIC_MEAN has 685 records (35.58%) with missing values.
Variable BLOODPRESSURE_SISTOLIC_MEAN has 685 records (35.58%) with missing values.
Variable HEART_RATE_MEAN has 685 records (35.58%) with missing values.
Variable RESPIRATORY_RATE_MEAN has 748 records (38.86%) with missing values.
Variable TEMPERATURE_MEAN has 694 records (36.05%) with missing values.
Variable OXYGEN_SATURATION_MEAN has 686 records (35.64%) with missing values.
Variable BLOODPRESSURE_DIASTOLIC_MEDIAN has 685 records (35.58%) with missing values.
Variable BLOODPRESSURE_SISTOLIC_MEDIAN has 685 records (35.58%) with missing values.
Variable HEART_RATE_MEDIAN has 685 records (35.58%) with missing values.
Variable RESPIRATORY_RATE_MEDIAN has 748 records (38.86%) with missing values.
Variable TEMPERATURE_MEDIAN has 694 records (36.05%) with missing values.
Variable OXYGEN_SATURATION_MEDIAN has 686 records (35.64%) with missing values.
Variable BLOODPRESSURE_DIASTOLIC_MIN has 685 records (35.58%) with missing values.
Variable BLOODPRESSURE_SISTOLIC_MIN has 685 records (35.58%) with missing values.
Variable HEART_RATE_MIN has 685 records (35.58%) with missing values.
Variable RESPIRATORY_RATE_MIN has 748 records (38.86%) with missing values.
Variable TEMPERATURE_MIN has 694 records (36.05%) with missing values.
Variable OXYGEN_SATURATION_MIN has 686 records (35.64%) with missing values.
Variable BLOODPRESSURE_DIASTOLIC_MAX has 685 records (35.58%) with missing values.
Variable BLOODPRESSURE_SISTOLIC_MAX has 685 records (35.58%) with missing values.
Variable HEART_RATE_MAX has 685 records (35.58%) with missing values.
Variable RESPIRATORY_RATE_MAX has 748 records (38.86%) with missing values.
Variable TEMPERATURE_MAX has 694 records (36.05%) with missing values.
Variable OXYGEN_SATURATION_MAX has 686 records (35.64%) with missing values.
Variable BLOODPRESSURE_DIASTOLIC_DIFF has 685 records (35.58%) with missing values.
Variable BLOODPRESSURE_SISTOLIC_DIFF has 685 records (35.58%) with missing values.
Variable HEART_RATE_DIFF has 685 records (35.58%) with missing values.
Variable RESPIRATORY_RATE_DIFF has 748 records (38.86%) with missing values.
Variable TEMPERATURE_DIFF has 694 records (36.05%) with missing values.
Variable OXYGEN_SATURATION_DIFF has 686 records (35.64%) with missing values.
Variable BLOODPRESSURE_DIASTOLIC_DIFF_REL has 685 records (35.58%) with missing values.
Variable BLOODPRESSURE_SISTOLIC_DIFF_REL has 685 records (35.58%) with missing values.
Variable HEART_RATE_DIFF_REL has 685 records (35.58%) with missing values.
Variable RESPIRATORY_RATE_DIFF_REL has 748 records (38.86%) with missing values.
Variable TEMPERATURE_DIFF_REL has 694 records (36.05%) with missing values.
Variable OXYGEN_SATURATION_DIFF_REL has 686 records (35.64%) with missing values.
In total, there are 225 variables with missing values


df_admitted = (raw_data.groupby("PATIENT_VISIT_IDENTIFIER")["ICU"].sum()>0).reset_index()*1
df_admitted.columns = ["PATIENT_VISIT_IDENTIFIER", "GO_ICU"]


raw_data = pd.merge(raw_data, df_admitted, on = "PATIENT_VISIT_IDENTIFIER")


# Create dataframe with the same colums as in raw_data
data = pd.DataFrame(columns=raw_data.columns)

# Create empty list which will hold data of each patient (mean)
patient_first_window = []

# for loop which goes through all patient id 
for patient_id in range(0, number_of_patients):

    # Create new dataframe which contains ONLY one patient data
    one_patient_data = raw_data.loc[raw_data['PATIENT_VISIT_IDENTIFIER'] == patient_id]
    
    # Fill up missing data with mean of available data (available data == particaular patient data)
    one_patient_data.fillna(one_patient_data.mean(),inplace=True)
    
    # using dataframe.iloc[[0]] => returns a dataframe
    patient_first_window.append(one_patient_data.iloc[[0]])
    
# Create new DataFrame with mean data of each patient cobined into one row (window 0-2)
# Add patient to the new main Dataframe which will contain all patients data
data = pd.concat(patient_first_window)

/var/folders/94/vt0_m_ds4ldfdtdt98l0j_cr0000gn/T/ipykernel_83979/2925090691.py:14: FutureWarning:

The default value of numeric_only in DataFrame.mean is deprecated. In a future version, it will default to False. In addition, specifying 'numeric_only=None' is deprecated. Select only valid columns or specify the value of numeric_only to silence this warning.

/var/folders/94/vt0_m_ds4ldfdtdt98l0j_cr0000gn/T/ipykernel_83979/2925090691.py:14: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy


data.shape

(385, 232)


data.head()


#drop rows with ICU == 1 ie drop data when the target variable is present, as stipulated by dataset author
data = data[data.ICU == 0].reset_index(drop = True)


data.shape

(353, 232)


#drop unnecessary columns
data = data.drop(["WINDOW", "ICU"],axis = 1)


data.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 353 entries, 0 to 352
Columns: 230 entries, PATIENT_VISIT_IDENTIFIER to GO_ICU
dtypes: float64(225), int64(4), object(1)
memory usage: 634.4+ KB


objects = data.select_dtypes(object).columns
print(objects)

Index(['AGE_PERCENTIL'], dtype='object')


data['PATIENT_VISIT_IDENTIFIER'] = data['PATIENT_VISIT_IDENTIFIER'].astype('int64')
data['AGE_ABOVE65'] = data['AGE_ABOVE65'].astype('int64')
data['GENDER'] = data['GENDER'].astype('int64')


raw_data.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 1925 entries, 0 to 1924
Columns: 232 entries, PATIENT_VISIT_IDENTIFIER to GO_ICU
dtypes: float64(225), int64(5), object(2)
memory usage: 3.4+ MB


data.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 353 entries, 0 to 352
Columns: 230 entries, PATIENT_VISIT_IDENTIFIER to GO_ICU
dtypes: float64(225), int64(4), object(1)
memory usage: 634.4+ KB


def check_for_missing_data_in_columns(dataset_name):
    print('NaN values =', dataset_name.isnull().sum().sum())
    print("""""")

    vars_with_missing = []

    for feature in dataset_name.columns:
        missings = dataset_name[feature].isna().sum()

        if missings > 0 :
            vars_with_missing.append(feature)
            missings_perc = missings / dataset_name.shape[0]

            print('Column {} has {} records ({:.2%}) with missing values.'.format(feature, missings, missings_perc))
    print('In total, there are {} columns with missing values'.format(len(vars_with_missing)))
    
check_for_missing_data_in_columns(data)

NaN values = 405

Column DISEASE GROUPING 1 has 1 records (0.28%) with missing values.
Column DISEASE GROUPING 2 has 1 records (0.28%) with missing values.
Column DISEASE GROUPING 3 has 1 records (0.28%) with missing values.
Column DISEASE GROUPING 4 has 1 records (0.28%) with missing values.
Column DISEASE GROUPING 5 has 1 records (0.28%) with missing values.
Column DISEASE GROUPING 6 has 1 records (0.28%) with missing values.
Column HTN has 1 records (0.28%) with missing values.
Column IMMUNOCOMPROMISED has 1 records (0.28%) with missing values.
Column OTHER has 1 records (0.28%) with missing values.
Column ALBUMIN_MEDIAN has 2 records (0.57%) with missing values.
Column ALBUMIN_MEAN has 2 records (0.57%) with missing values.
Column ALBUMIN_MIN has 2 records (0.57%) with missing values.
Column ALBUMIN_MAX has 2 records (0.57%) with missing values.
Column ALBUMIN_DIFF has 2 records (0.57%) with missing values.
Column BE_ARTERIAL_MEDIAN has 2 records (0.57%) with missing values.
Column BE_ARTERIAL_MEAN has 2 records (0.57%) with missing values.
Column BE_ARTERIAL_MIN has 2 records (0.57%) with missing values.
Column BE_ARTERIAL_MAX has 2 records (0.57%) with missing values.
Column BE_ARTERIAL_DIFF has 2 records (0.57%) with missing values.
Column BE_VENOUS_MEDIAN has 2 records (0.57%) with missing values.
Column BE_VENOUS_MEAN has 2 records (0.57%) with missing values.
Column BE_VENOUS_MIN has 2 records (0.57%) with missing values.
Column BE_VENOUS_MAX has 2 records (0.57%) with missing values.
Column BE_VENOUS_DIFF has 2 records (0.57%) with missing values.
Column BIC_ARTERIAL_MEDIAN has 2 records (0.57%) with missing values.
Column BIC_ARTERIAL_MEAN has 2 records (0.57%) with missing values.
Column BIC_ARTERIAL_MIN has 2 records (0.57%) with missing values.
Column BIC_ARTERIAL_MAX has 2 records (0.57%) with missing values.
Column BIC_ARTERIAL_DIFF has 2 records (0.57%) with missing values.
Column BIC_VENOUS_MEDIAN has 2 records (0.57%) with missing values.
Column BIC_VENOUS_MEAN has 2 records (0.57%) with missing values.
Column BIC_VENOUS_MIN has 2 records (0.57%) with missing values.
Column BIC_VENOUS_MAX has 2 records (0.57%) with missing values.
Column BIC_VENOUS_DIFF has 2 records (0.57%) with missing values.
Column BILLIRUBIN_MEDIAN has 2 records (0.57%) with missing values.
Column BILLIRUBIN_MEAN has 2 records (0.57%) with missing values.
Column BILLIRUBIN_MIN has 2 records (0.57%) with missing values.
Column BILLIRUBIN_MAX has 2 records (0.57%) with missing values.
Column BILLIRUBIN_DIFF has 2 records (0.57%) with missing values.
Column BLAST_MEDIAN has 2 records (0.57%) with missing values.
Column BLAST_MEAN has 2 records (0.57%) with missing values.
Column BLAST_MIN has 2 records (0.57%) with missing values.
Column BLAST_MAX has 2 records (0.57%) with missing values.
Column BLAST_DIFF has 2 records (0.57%) with missing values.
Column CALCIUM_MEDIAN has 2 records (0.57%) with missing values.
Column CALCIUM_MEAN has 2 records (0.57%) with missing values.
Column CALCIUM_MIN has 2 records (0.57%) with missing values.
Column CALCIUM_MAX has 2 records (0.57%) with missing values.
Column CALCIUM_DIFF has 2 records (0.57%) with missing values.
Column CREATININ_MEDIAN has 2 records (0.57%) with missing values.
Column CREATININ_MEAN has 2 records (0.57%) with missing values.
Column CREATININ_MIN has 2 records (0.57%) with missing values.
Column CREATININ_MAX has 2 records (0.57%) with missing values.
Column CREATININ_DIFF has 2 records (0.57%) with missing values.
Column FFA_MEDIAN has 2 records (0.57%) with missing values.
Column FFA_MEAN has 2 records (0.57%) with missing values.
Column FFA_MIN has 2 records (0.57%) with missing values.
Column FFA_MAX has 2 records (0.57%) with missing values.
Column FFA_DIFF has 2 records (0.57%) with missing values.
Column GGT_MEDIAN has 2 records (0.57%) with missing values.
Column GGT_MEAN has 2 records (0.57%) with missing values.
Column GGT_MIN has 2 records (0.57%) with missing values.
Column GGT_MAX has 2 records (0.57%) with missing values.
Column GGT_DIFF has 2 records (0.57%) with missing values.
Column GLUCOSE_MEDIAN has 2 records (0.57%) with missing values.
Column GLUCOSE_MEAN has 2 records (0.57%) with missing values.
Column GLUCOSE_MIN has 2 records (0.57%) with missing values.
Column GLUCOSE_MAX has 2 records (0.57%) with missing values.
Column GLUCOSE_DIFF has 2 records (0.57%) with missing values.
Column HEMATOCRITE_MEDIAN has 2 records (0.57%) with missing values.
Column HEMATOCRITE_MEAN has 2 records (0.57%) with missing values.
Column HEMATOCRITE_MIN has 2 records (0.57%) with missing values.
Column HEMATOCRITE_MAX has 2 records (0.57%) with missing values.
Column HEMATOCRITE_DIFF has 2 records (0.57%) with missing values.
Column HEMOGLOBIN_MEDIAN has 2 records (0.57%) with missing values.
Column HEMOGLOBIN_MEAN has 2 records (0.57%) with missing values.
Column HEMOGLOBIN_MIN has 2 records (0.57%) with missing values.
Column HEMOGLOBIN_MAX has 2 records (0.57%) with missing values.
Column HEMOGLOBIN_DIFF has 2 records (0.57%) with missing values.
Column INR_MEDIAN has 2 records (0.57%) with missing values.
Column INR_MEAN has 2 records (0.57%) with missing values.
Column INR_MIN has 2 records (0.57%) with missing values.
Column INR_MAX has 2 records (0.57%) with missing values.
Column INR_DIFF has 2 records (0.57%) with missing values.
Column LACTATE_MEDIAN has 2 records (0.57%) with missing values.
Column LACTATE_MEAN has 2 records (0.57%) with missing values.
Column LACTATE_MIN has 2 records (0.57%) with missing values.
Column LACTATE_MAX has 2 records (0.57%) with missing values.
Column LACTATE_DIFF has 2 records (0.57%) with missing values.
Column LEUKOCYTES_MEDIAN has 2 records (0.57%) with missing values.
Column LEUKOCYTES_MEAN has 2 records (0.57%) with missing values.
Column LEUKOCYTES_MIN has 2 records (0.57%) with missing values.
Column LEUKOCYTES_MAX has 2 records (0.57%) with missing values.
Column LEUKOCYTES_DIFF has 2 records (0.57%) with missing values.
Column LINFOCITOS_MEDIAN has 2 records (0.57%) with missing values.
Column LINFOCITOS_MEAN has 2 records (0.57%) with missing values.
Column LINFOCITOS_MIN has 2 records (0.57%) with missing values.
Column LINFOCITOS_MAX has 2 records (0.57%) with missing values.
Column LINFOCITOS_DIFF has 2 records (0.57%) with missing values.
Column NEUTROPHILES_MEDIAN has 2 records (0.57%) with missing values.
Column NEUTROPHILES_MEAN has 2 records (0.57%) with missing values.
Column NEUTROPHILES_MIN has 2 records (0.57%) with missing values.
Column NEUTROPHILES_MAX has 2 records (0.57%) with missing values.
Column NEUTROPHILES_DIFF has 2 records (0.57%) with missing values.
Column P02_ARTERIAL_MEDIAN has 2 records (0.57%) with missing values.
Column P02_ARTERIAL_MEAN has 2 records (0.57%) with missing values.
Column P02_ARTERIAL_MIN has 2 records (0.57%) with missing values.
Column P02_ARTERIAL_MAX has 2 records (0.57%) with missing values.
Column P02_ARTERIAL_DIFF has 2 records (0.57%) with missing values.
Column P02_VENOUS_MEDIAN has 2 records (0.57%) with missing values.
Column P02_VENOUS_MEAN has 2 records (0.57%) with missing values.
Column P02_VENOUS_MIN has 2 records (0.57%) with missing values.
Column P02_VENOUS_MAX has 2 records (0.57%) with missing values.
Column P02_VENOUS_DIFF has 2 records (0.57%) with missing values.
Column PC02_ARTERIAL_MEDIAN has 2 records (0.57%) with missing values.
Column PC02_ARTERIAL_MEAN has 2 records (0.57%) with missing values.
Column PC02_ARTERIAL_MIN has 2 records (0.57%) with missing values.
Column PC02_ARTERIAL_MAX has 2 records (0.57%) with missing values.
Column PC02_ARTERIAL_DIFF has 2 records (0.57%) with missing values.
Column PC02_VENOUS_MEDIAN has 2 records (0.57%) with missing values.
Column PC02_VENOUS_MEAN has 2 records (0.57%) with missing values.
Column PC02_VENOUS_MIN has 2 records (0.57%) with missing values.
Column PC02_VENOUS_MAX has 2 records (0.57%) with missing values.
Column PC02_VENOUS_DIFF has 2 records (0.57%) with missing values.
Column PCR_MEDIAN has 2 records (0.57%) with missing values.
Column PCR_MEAN has 2 records (0.57%) with missing values.
Column PCR_MIN has 2 records (0.57%) with missing values.
Column PCR_MAX has 2 records (0.57%) with missing values.
Column PCR_DIFF has 2 records (0.57%) with missing values.
Column PH_ARTERIAL_MEDIAN has 2 records (0.57%) with missing values.
Column PH_ARTERIAL_MEAN has 2 records (0.57%) with missing values.
Column PH_ARTERIAL_MIN has 2 records (0.57%) with missing values.
Column PH_ARTERIAL_MAX has 2 records (0.57%) with missing values.
Column PH_ARTERIAL_DIFF has 2 records (0.57%) with missing values.
Column PH_VENOUS_MEDIAN has 2 records (0.57%) with missing values.
Column PH_VENOUS_MEAN has 2 records (0.57%) with missing values.
Column PH_VENOUS_MIN has 2 records (0.57%) with missing values.
Column PH_VENOUS_MAX has 2 records (0.57%) with missing values.
Column PH_VENOUS_DIFF has 2 records (0.57%) with missing values.
Column PLATELETS_MEDIAN has 2 records (0.57%) with missing values.
Column PLATELETS_MEAN has 2 records (0.57%) with missing values.
Column PLATELETS_MIN has 2 records (0.57%) with missing values.
Column PLATELETS_MAX has 2 records (0.57%) with missing values.
Column PLATELETS_DIFF has 2 records (0.57%) with missing values.
Column POTASSIUM_MEDIAN has 2 records (0.57%) with missing values.
Column POTASSIUM_MEAN has 2 records (0.57%) with missing values.
Column POTASSIUM_MIN has 2 records (0.57%) with missing values.
Column POTASSIUM_MAX has 2 records (0.57%) with missing values.
Column POTASSIUM_DIFF has 2 records (0.57%) with missing values.
Column SAT02_ARTERIAL_MEDIAN has 2 records (0.57%) with missing values.
Column SAT02_ARTERIAL_MEAN has 2 records (0.57%) with missing values.
Column SAT02_ARTERIAL_MIN has 2 records (0.57%) with missing values.
Column SAT02_ARTERIAL_MAX has 2 records (0.57%) with missing values.
Column SAT02_ARTERIAL_DIFF has 2 records (0.57%) with missing values.
Column SAT02_VENOUS_MEDIAN has 2 records (0.57%) with missing values.
Column SAT02_VENOUS_MEAN has 2 records (0.57%) with missing values.
Column SAT02_VENOUS_MIN has 2 records (0.57%) with missing values.
Column SAT02_VENOUS_MAX has 2 records (0.57%) with missing values.
Column SAT02_VENOUS_DIFF has 2 records (0.57%) with missing values.
Column SODIUM_MEDIAN has 2 records (0.57%) with missing values.
Column SODIUM_MEAN has 2 records (0.57%) with missing values.
Column SODIUM_MIN has 2 records (0.57%) with missing values.
Column SODIUM_MAX has 2 records (0.57%) with missing values.
Column SODIUM_DIFF has 2 records (0.57%) with missing values.
Column TGO_MEDIAN has 2 records (0.57%) with missing values.
Column TGO_MEAN has 2 records (0.57%) with missing values.
Column TGO_MIN has 2 records (0.57%) with missing values.
Column TGO_MAX has 2 records (0.57%) with missing values.
Column TGO_DIFF has 2 records (0.57%) with missing values.
Column TGP_MEDIAN has 2 records (0.57%) with missing values.
Column TGP_MEAN has 2 records (0.57%) with missing values.
Column TGP_MIN has 2 records (0.57%) with missing values.
Column TGP_MAX has 2 records (0.57%) with missing values.
Column TGP_DIFF has 2 records (0.57%) with missing values.
Column TTPA_MEDIAN has 2 records (0.57%) with missing values.
Column TTPA_MEAN has 2 records (0.57%) with missing values.
Column TTPA_MIN has 2 records (0.57%) with missing values.
Column TTPA_MAX has 2 records (0.57%) with missing values.
Column TTPA_DIFF has 2 records (0.57%) with missing values.
Column UREA_MEDIAN has 2 records (0.57%) with missing values.
Column UREA_MEAN has 2 records (0.57%) with missing values.
Column UREA_MIN has 2 records (0.57%) with missing values.
Column UREA_MAX has 2 records (0.57%) with missing values.
Column UREA_DIFF has 2 records (0.57%) with missing values.
Column DIMER_MEDIAN has 2 records (0.57%) with missing values.
Column DIMER_MEAN has 2 records (0.57%) with missing values.
Column DIMER_MIN has 2 records (0.57%) with missing values.
Column DIMER_MAX has 2 records (0.57%) with missing values.
Column DIMER_DIFF has 2 records (0.57%) with missing values.
Column BLOODPRESSURE_DIASTOLIC_MEAN has 1 records (0.28%) with missing values.
Column BLOODPRESSURE_SISTOLIC_MEAN has 1 records (0.28%) with missing values.
Column HEART_RATE_MEAN has 1 records (0.28%) with missing values.
Column RESPIRATORY_RATE_MEAN has 1 records (0.28%) with missing values.
Column TEMPERATURE_MEAN has 1 records (0.28%) with missing values.
Column OXYGEN_SATURATION_MEAN has 1 records (0.28%) with missing values.
Column BLOODPRESSURE_DIASTOLIC_MEDIAN has 1 records (0.28%) with missing values.
Column BLOODPRESSURE_SISTOLIC_MEDIAN has 1 records (0.28%) with missing values.
Column HEART_RATE_MEDIAN has 1 records (0.28%) with missing values.
Column RESPIRATORY_RATE_MEDIAN has 1 records (0.28%) with missing values.
Column TEMPERATURE_MEDIAN has 1 records (0.28%) with missing values.
Column OXYGEN_SATURATION_MEDIAN has 1 records (0.28%) with missing values.
Column BLOODPRESSURE_DIASTOLIC_MIN has 1 records (0.28%) with missing values.
Column BLOODPRESSURE_SISTOLIC_MIN has 1 records (0.28%) with missing values.
Column HEART_RATE_MIN has 1 records (0.28%) with missing values.
Column RESPIRATORY_RATE_MIN has 1 records (0.28%) with missing values.
Column TEMPERATURE_MIN has 1 records (0.28%) with missing values.
Column OXYGEN_SATURATION_MIN has 1 records (0.28%) with missing values.
Column BLOODPRESSURE_DIASTOLIC_MAX has 1 records (0.28%) with missing values.
Column BLOODPRESSURE_SISTOLIC_MAX has 1 records (0.28%) with missing values.
Column HEART_RATE_MAX has 1 records (0.28%) with missing values.
Column RESPIRATORY_RATE_MAX has 1 records (0.28%) with missing values.
Column TEMPERATURE_MAX has 1 records (0.28%) with missing values.
Column OXYGEN_SATURATION_MAX has 1 records (0.28%) with missing values.
Column BLOODPRESSURE_DIASTOLIC_DIFF has 1 records (0.28%) with missing values.
Column BLOODPRESSURE_SISTOLIC_DIFF has 1 records (0.28%) with missing values.
Column HEART_RATE_DIFF has 1 records (0.28%) with missing values.
Column RESPIRATORY_RATE_DIFF has 1 records (0.28%) with missing values.
Column TEMPERATURE_DIFF has 1 records (0.28%) with missing values.
Column OXYGEN_SATURATION_DIFF has 1 records (0.28%) with missing values.
Column BLOODPRESSURE_DIASTOLIC_DIFF_REL has 1 records (0.28%) with missing values.
Column BLOODPRESSURE_SISTOLIC_DIFF_REL has 1 records (0.28%) with missing values.
Column HEART_RATE_DIFF_REL has 1 records (0.28%) with missing values.
Column RESPIRATORY_RATE_DIFF_REL has 1 records (0.28%) with missing values.
Column TEMPERATURE_DIFF_REL has 1 records (0.28%) with missing values.
Column OXYGEN_SATURATION_DIFF_REL has 1 records (0.28%) with missing values.
In total, there are 225 columns with missing values


empty_data = data[data.isna().any(axis=1)]
display(empty_data)


# drop rows where at least one value is NaN
dataset = data.dropna(axis=0)
dataset = dataset.reset_index(drop=True)


check_for_missing_data_in_columns(dataset)

NaN values = 0

In total, there are 0 columns with missing values


dataset.shape

(351, 230)


# check for duplicate columns i.e columns having all values exactly equal
duplicated = dataset.T.duplicated()
count = 0

# The first instance of a column is marked as 'False', when the value is seen to be repeated in other 
# column(s) it is marked as 'True'
for i in range(len(duplicated)):
    if duplicated[i] == True:
        print(duplicated.index[i], ":", duplicated[i])
        count += 1
        
# 143 duplicate columns
print("total duplicate columns = ", count)

ALBUMIN_MEAN : True
ALBUMIN_MIN : True
ALBUMIN_MAX : True
BE_ARTERIAL_MEAN : True
BE_ARTERIAL_MIN : True
BE_ARTERIAL_MAX : True
BE_ARTERIAL_DIFF : True
BE_VENOUS_MEAN : True
BE_VENOUS_MIN : True
BE_VENOUS_MAX : True
BE_VENOUS_DIFF : True
BIC_ARTERIAL_MEAN : True
BIC_ARTERIAL_MIN : True
BIC_ARTERIAL_MAX : True
BIC_ARTERIAL_DIFF : True
BIC_VENOUS_MEAN : True
BIC_VENOUS_MIN : True
BIC_VENOUS_MAX : True
BIC_VENOUS_DIFF : True
BILLIRUBIN_MEAN : True
BILLIRUBIN_MIN : True
BILLIRUBIN_MAX : True
BILLIRUBIN_DIFF : True
BLAST_MEAN : True
BLAST_MIN : True
BLAST_MAX : True
BLAST_DIFF : True
CALCIUM_MEAN : True
CALCIUM_MIN : True
CALCIUM_MAX : True
CALCIUM_DIFF : True
CREATININ_MEAN : True
CREATININ_MIN : True
CREATININ_MAX : True
CREATININ_DIFF : True
FFA_MEAN : True
FFA_MIN : True
FFA_MAX : True
FFA_DIFF : True
GGT_MEAN : True
GGT_MIN : True
GGT_MAX : True
GGT_DIFF : True
GLUCOSE_MEAN : True
GLUCOSE_MIN : True
GLUCOSE_MAX : True
GLUCOSE_DIFF : True
HEMATOCRITE_MEAN : True
HEMATOCRITE_MIN : True
HEMATOCRITE_MAX : True
HEMATOCRITE_DIFF : True
HEMOGLOBIN_MEAN : True
HEMOGLOBIN_MIN : True
HEMOGLOBIN_MAX : True
HEMOGLOBIN_DIFF : True
INR_MEAN : True
INR_MIN : True
INR_MAX : True
INR_DIFF : True
LACTATE_MEAN : True
LACTATE_MIN : True
LACTATE_MAX : True
LACTATE_DIFF : True
LEUKOCYTES_MEAN : True
LEUKOCYTES_MIN : True
LEUKOCYTES_MAX : True
LEUKOCYTES_DIFF : True
LINFOCITOS_MEAN : True
LINFOCITOS_MIN : True
LINFOCITOS_MAX : True
LINFOCITOS_DIFF : True
NEUTROPHILES_MEAN : True
NEUTROPHILES_MIN : True
NEUTROPHILES_MAX : True
NEUTROPHILES_DIFF : True
P02_ARTERIAL_MEAN : True
P02_ARTERIAL_MIN : True
P02_ARTERIAL_MAX : True
P02_ARTERIAL_DIFF : True
P02_VENOUS_MEAN : True
P02_VENOUS_MIN : True
P02_VENOUS_MAX : True
P02_VENOUS_DIFF : True
PC02_ARTERIAL_MEAN : True
PC02_ARTERIAL_MIN : True
PC02_ARTERIAL_MAX : True
PC02_ARTERIAL_DIFF : True
PC02_VENOUS_MEAN : True
PC02_VENOUS_MIN : True
PC02_VENOUS_MAX : True
PC02_VENOUS_DIFF : True
PCR_MEAN : True
PCR_MIN : True
PCR_MAX : True
PCR_DIFF : True
PH_ARTERIAL_MEAN : True
PH_ARTERIAL_MIN : True
PH_ARTERIAL_MAX : True
PH_ARTERIAL_DIFF : True
PH_VENOUS_MEAN : True
PH_VENOUS_MIN : True
PH_VENOUS_MAX : True
PH_VENOUS_DIFF : True
PLATELETS_MEAN : True
PLATELETS_MIN : True
PLATELETS_MAX : True
PLATELETS_DIFF : True
POTASSIUM_MEAN : True
POTASSIUM_MIN : True
POTASSIUM_MAX : True
POTASSIUM_DIFF : True
SAT02_ARTERIAL_MEAN : True
SAT02_ARTERIAL_MIN : True
SAT02_ARTERIAL_MAX : True
SAT02_ARTERIAL_DIFF : True
SAT02_VENOUS_MEAN : True
SAT02_VENOUS_MIN : True
SAT02_VENOUS_MAX : True
SAT02_VENOUS_DIFF : True
SODIUM_MEAN : True
SODIUM_MIN : True
SODIUM_MAX : True
SODIUM_DIFF : True
TGO_MEAN : True
TGO_MIN : True
TGO_MAX : True
TGO_DIFF : True
TGP_MEAN : True
TGP_MIN : True
TGP_MAX : True
TGP_DIFF : True
TTPA_MEAN : True
TTPA_MIN : True
TTPA_MAX : True
TTPA_DIFF : True
UREA_MEAN : True
UREA_MIN : True
UREA_MAX : True
UREA_DIFF : True
DIMER_MEAN : True
DIMER_MIN : True
DIMER_MAX : True
DIMER_DIFF : True
total duplicate columns =  143


dataset = dataset.T.drop_duplicates().T


# Drop columns with constant value
col = dataset.columns[dataset.nunique() <= 1].tolist()

print(col)

dataset.drop(columns = col, inplace=True)

['ALBUMIN_DIFF']


dataset.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 351 entries, 0 to 350
Data columns (total 86 columns):
 #   Column                            Non-Null Count  Dtype 
---  ------                            --------------  ----- 
 0   PATIENT_VISIT_IDENTIFIER          351 non-null    object
 1   AGE_ABOVE65                       351 non-null    object
 2   AGE_PERCENTIL                     351 non-null    object
 3   GENDER                            351 non-null    object
 4   DISEASE GROUPING 1                351 non-null    object
 5   DISEASE GROUPING 2                351 non-null    object
 6   DISEASE GROUPING 3                351 non-null    object
 7   DISEASE GROUPING 4                351 non-null    object
 8   DISEASE GROUPING 5                351 non-null    object
 9   DISEASE GROUPING 6                351 non-null    object
 10  HTN                               351 non-null    object
 11  IMMUNOCOMPROMISED                 351 non-null    object
 12  OTHER                             351 non-null    object
 13  ALBUMIN_MEDIAN                    351 non-null    object
 14  BE_ARTERIAL_MEDIAN                351 non-null    object
 15  BE_VENOUS_MEDIAN                  351 non-null    object
 16  BIC_ARTERIAL_MEDIAN               351 non-null    object
 17  BIC_VENOUS_MEDIAN                 351 non-null    object
 18  BILLIRUBIN_MEDIAN                 351 non-null    object
 19  BLAST_MEDIAN                      351 non-null    object
 20  CALCIUM_MEDIAN                    351 non-null    object
 21  CREATININ_MEDIAN                  351 non-null    object
 22  FFA_MEDIAN                        351 non-null    object
 23  GGT_MEDIAN                        351 non-null    object
 24  GLUCOSE_MEDIAN                    351 non-null    object
 25  HEMATOCRITE_MEDIAN                351 non-null    object
 26  HEMOGLOBIN_MEDIAN                 351 non-null    object
 27  INR_MEDIAN                        351 non-null    object
 28  LACTATE_MEDIAN                    351 non-null    object
 29  LEUKOCYTES_MEDIAN                 351 non-null    object
 30  LINFOCITOS_MEDIAN                 351 non-null    object
 31  NEUTROPHILES_MEDIAN               351 non-null    object
 32  P02_ARTERIAL_MEDIAN               351 non-null    object
 33  P02_VENOUS_MEDIAN                 351 non-null    object
 34  PC02_ARTERIAL_MEDIAN              351 non-null    object
 35  PC02_VENOUS_MEDIAN                351 non-null    object
 36  PCR_MEDIAN                        351 non-null    object
 37  PH_ARTERIAL_MEDIAN                351 non-null    object
 38  PH_VENOUS_MEDIAN                  351 non-null    object
 39  PLATELETS_MEDIAN                  351 non-null    object
 40  POTASSIUM_MEDIAN                  351 non-null    object
 41  SAT02_ARTERIAL_MEDIAN             351 non-null    object
 42  SAT02_VENOUS_MEDIAN               351 non-null    object
 43  SODIUM_MEDIAN                     351 non-null    object
 44  TGO_MEDIAN                        351 non-null    object
 45  TGP_MEDIAN                        351 non-null    object
 46  TTPA_MEDIAN                       351 non-null    object
 47  UREA_MEDIAN                       351 non-null    object
 48  DIMER_MEDIAN                      351 non-null    object
 49  BLOODPRESSURE_DIASTOLIC_MEAN      351 non-null    object
 50  BLOODPRESSURE_SISTOLIC_MEAN       351 non-null    object
 51  HEART_RATE_MEAN                   351 non-null    object
 52  RESPIRATORY_RATE_MEAN             351 non-null    object
 53  TEMPERATURE_MEAN                  351 non-null    object
 54  OXYGEN_SATURATION_MEAN            351 non-null    object
 55  BLOODPRESSURE_DIASTOLIC_MEDIAN    351 non-null    object
 56  BLOODPRESSURE_SISTOLIC_MEDIAN     351 non-null    object
 57  HEART_RATE_MEDIAN                 351 non-null    object
 58  RESPIRATORY_RATE_MEDIAN           351 non-null    object
 59  TEMPERATURE_MEDIAN                351 non-null    object
 60  OXYGEN_SATURATION_MEDIAN          351 non-null    object
 61  BLOODPRESSURE_DIASTOLIC_MIN       351 non-null    object
 62  BLOODPRESSURE_SISTOLIC_MIN        351 non-null    object
 63  HEART_RATE_MIN                    351 non-null    object
 64  RESPIRATORY_RATE_MIN              351 non-null    object
 65  TEMPERATURE_MIN                   351 non-null    object
 66  OXYGEN_SATURATION_MIN             351 non-null    object
 67  BLOODPRESSURE_DIASTOLIC_MAX       351 non-null    object
 68  BLOODPRESSURE_SISTOLIC_MAX        351 non-null    object
 69  HEART_RATE_MAX                    351 non-null    object
 70  RESPIRATORY_RATE_MAX              351 non-null    object
 71  TEMPERATURE_MAX                   351 non-null    object
 72  OXYGEN_SATURATION_MAX             351 non-null    object
 73  BLOODPRESSURE_DIASTOLIC_DIFF      351 non-null    object
 74  BLOODPRESSURE_SISTOLIC_DIFF       351 non-null    object
 75  HEART_RATE_DIFF                   351 non-null    object
 76  RESPIRATORY_RATE_DIFF             351 non-null    object
 77  TEMPERATURE_DIFF                  351 non-null    object
 78  OXYGEN_SATURATION_DIFF            351 non-null    object
 79  BLOODPRESSURE_DIASTOLIC_DIFF_REL  351 non-null    object
 80  BLOODPRESSURE_SISTOLIC_DIFF_REL   351 non-null    object
 81  HEART_RATE_DIFF_REL               351 non-null    object
 82  RESPIRATORY_RATE_DIFF_REL         351 non-null    object
 83  TEMPERATURE_DIFF_REL              351 non-null    object
 84  OXYGEN_SATURATION_DIFF_REL        351 non-null    object
 85  GO_ICU                            351 non-null    object
dtypes: object(86)
memory usage: 236.0+ KB


dataset.shape

(351, 86)


for column in dataset.columns:
    try:
        dataset[column] = dataset[column].astype("float64")
    except ValueError:
        pass


dataset['AGE_ABOVE65'] = dataset['AGE_ABOVE65'].astype('int64')
dataset['GENDER'] = dataset['GENDER'].astype('int64')
dataset['PATIENT_VISIT_IDENTIFIER'] = dataset['PATIENT_VISIT_IDENTIFIER'].astype('int64')
dataset['GO_ICU'] = dataset['GO_ICU'].astype('int64')


dataset.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 351 entries, 0 to 350
Data columns (total 86 columns):
 #   Column                            Non-Null Count  Dtype  
---  ------                            --------------  -----  
 0   PATIENT_VISIT_IDENTIFIER          351 non-null    int64  
 1   AGE_ABOVE65                       351 non-null    int64  
 2   AGE_PERCENTIL                     351 non-null    object 
 3   GENDER                            351 non-null    int64  
 4   DISEASE GROUPING 1                351 non-null    float64
 5   DISEASE GROUPING 2                351 non-null    float64
 6   DISEASE GROUPING 3                351 non-null    float64
 7   DISEASE GROUPING 4                351 non-null    float64
 8   DISEASE GROUPING 5                351 non-null    float64
 9   DISEASE GROUPING 6                351 non-null    float64
 10  HTN                               351 non-null    float64
 11  IMMUNOCOMPROMISED                 351 non-null    float64
 12  OTHER                             351 non-null    float64
 13  ALBUMIN_MEDIAN                    351 non-null    float64
 14  BE_ARTERIAL_MEDIAN                351 non-null    float64
 15  BE_VENOUS_MEDIAN                  351 non-null    float64
 16  BIC_ARTERIAL_MEDIAN               351 non-null    float64
 17  BIC_VENOUS_MEDIAN                 351 non-null    float64
 18  BILLIRUBIN_MEDIAN                 351 non-null    float64
 19  BLAST_MEDIAN                      351 non-null    float64
 20  CALCIUM_MEDIAN                    351 non-null    float64
 21  CREATININ_MEDIAN                  351 non-null    float64
 22  FFA_MEDIAN                        351 non-null    float64
 23  GGT_MEDIAN                        351 non-null    float64
 24  GLUCOSE_MEDIAN                    351 non-null    float64
 25  HEMATOCRITE_MEDIAN                351 non-null    float64
 26  HEMOGLOBIN_MEDIAN                 351 non-null    float64
 27  INR_MEDIAN                        351 non-null    float64
 28  LACTATE_MEDIAN                    351 non-null    float64
 29  LEUKOCYTES_MEDIAN                 351 non-null    float64
 30  LINFOCITOS_MEDIAN                 351 non-null    float64
 31  NEUTROPHILES_MEDIAN               351 non-null    float64
 32  P02_ARTERIAL_MEDIAN               351 non-null    float64
 33  P02_VENOUS_MEDIAN                 351 non-null    float64
 34  PC02_ARTERIAL_MEDIAN              351 non-null    float64
 35  PC02_VENOUS_MEDIAN                351 non-null    float64
 36  PCR_MEDIAN                        351 non-null    float64
 37  PH_ARTERIAL_MEDIAN                351 non-null    float64
 38  PH_VENOUS_MEDIAN                  351 non-null    float64
 39  PLATELETS_MEDIAN                  351 non-null    float64
 40  POTASSIUM_MEDIAN                  351 non-null    float64
 41  SAT02_ARTERIAL_MEDIAN             351 non-null    float64
 42  SAT02_VENOUS_MEDIAN               351 non-null    float64
 43  SODIUM_MEDIAN                     351 non-null    float64
 44  TGO_MEDIAN                        351 non-null    float64
 45  TGP_MEDIAN                        351 non-null    float64
 46  TTPA_MEDIAN                       351 non-null    float64
 47  UREA_MEDIAN                       351 non-null    float64
 48  DIMER_MEDIAN                      351 non-null    float64
 49  BLOODPRESSURE_DIASTOLIC_MEAN      351 non-null    float64
 50  BLOODPRESSURE_SISTOLIC_MEAN       351 non-null    float64
 51  HEART_RATE_MEAN                   351 non-null    float64
 52  RESPIRATORY_RATE_MEAN             351 non-null    float64
 53  TEMPERATURE_MEAN                  351 non-null    float64
 54  OXYGEN_SATURATION_MEAN            351 non-null    float64
 55  BLOODPRESSURE_DIASTOLIC_MEDIAN    351 non-null    float64
 56  BLOODPRESSURE_SISTOLIC_MEDIAN     351 non-null    float64
 57  HEART_RATE_MEDIAN                 351 non-null    float64
 58  RESPIRATORY_RATE_MEDIAN           351 non-null    float64
 59  TEMPERATURE_MEDIAN                351 non-null    float64
 60  OXYGEN_SATURATION_MEDIAN          351 non-null    float64
 61  BLOODPRESSURE_DIASTOLIC_MIN       351 non-null    float64
 62  BLOODPRESSURE_SISTOLIC_MIN        351 non-null    float64
 63  HEART_RATE_MIN                    351 non-null    float64
 64  RESPIRATORY_RATE_MIN              351 non-null    float64
 65  TEMPERATURE_MIN                   351 non-null    float64
 66  OXYGEN_SATURATION_MIN             351 non-null    float64
 67  BLOODPRESSURE_DIASTOLIC_MAX       351 non-null    float64
 68  BLOODPRESSURE_SISTOLIC_MAX        351 non-null    float64
 69  HEART_RATE_MAX                    351 non-null    float64
 70  RESPIRATORY_RATE_MAX              351 non-null    float64
 71  TEMPERATURE_MAX                   351 non-null    float64
 72  OXYGEN_SATURATION_MAX             351 non-null    float64
 73  BLOODPRESSURE_DIASTOLIC_DIFF      351 non-null    float64
 74  BLOODPRESSURE_SISTOLIC_DIFF       351 non-null    float64
 75  HEART_RATE_DIFF                   351 non-null    float64
 76  RESPIRATORY_RATE_DIFF             351 non-null    float64
 77  TEMPERATURE_DIFF                  351 non-null    float64
 78  OXYGEN_SATURATION_DIFF            351 non-null    float64
 79  BLOODPRESSURE_DIASTOLIC_DIFF_REL  351 non-null    float64
 80  BLOODPRESSURE_SISTOLIC_DIFF_REL   351 non-null    float64
 81  HEART_RATE_DIFF_REL               351 non-null    float64
 82  RESPIRATORY_RATE_DIFF_REL         351 non-null    float64
 83  TEMPERATURE_DIFF_REL              351 non-null    float64
 84  OXYGEN_SATURATION_DIFF_REL        351 non-null    float64
 85  GO_ICU                            351 non-null    int64  
dtypes: float64(81), int64(4), object(1)
memory usage: 236.0+ KB


for num in range (1, 10):
    dataset["AGE_PERCENTIL"] = dataset["AGE_PERCENTIL"].replace([f'{num}0th'], f'{num}0')
dataset["AGE_PERCENTIL"] = dataset["AGE_PERCENTIL"].replace(['Above 90th'], '100')


dataset["AGE_PERCENTIL"] = dataset["AGE_PERCENTIL"].astype(int)
dataset["AGE_PERCENTIL"].info()

<class 'pandas.core.series.Series'>
RangeIndex: 351 entries, 0 to 350
Series name: AGE_PERCENTIL
Non-Null Count  Dtype
--------------  -----
351 non-null    int64
dtypes: int64(1)
memory usage: 2.9 KB


metadata = []
for feature in dataset.columns:
    # Defining the role
    if feature == 'GO_ICU':
        use = 'target'
    elif feature == 'PATIENT_VISIT_IDENTIFIER':
        use = 'patient id'
    else:
        use = 'input'
         
    # Defining the type
    if 'GROUPING' in feature or 'GENDER' in feature or 'AGE_ABOVE65' in feature or feature == 'GO_ICU':
        type = 'binary'
    elif 'AGE_PERCENTIL' in feature or feature == 'PATIENT_VISIT_IDENTIFIER':
        type = 'categorical'
    elif dataset[feature].dtype == float or isinstance(dataset[feature].dtype, float):
        type = 'real'
    else:
        type = 'integer'
        
    
    
    # Defining the data type 
    dtype = dataset[feature].dtype
    
    category = 'none'
    # Defining the category
    if 'DIFF' in feature:
        category = 'vital_signs'
    elif 'MAX' in feature:
        category = 'vital_signs'
    elif 'MIN' in feature:
        category = 'vital_signs'
    elif 'MEDIAN' in feature:
        category = 'vital_signs'
    elif 'MEAN' in feature:
        category = 'vital_signs'
    elif 'GROUPING' in feature:
        category = 'disease_group'
    
    
    # Creating a Dict that contains all the metadata for the variable
    feature_dictionary = {
        'varname': feature,
        'use': use,
        'type': type,
        'dtype': dtype,
        'category' : category
    }
    metadata.append(feature_dictionary)
    
meta = pd.DataFrame(metadata, columns=['varname', 'use', 'type', 'dtype', 'category'])
meta.set_index('varname', inplace=True)
meta


dataset.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 351 entries, 0 to 350
Data columns (total 86 columns):
 #   Column                            Non-Null Count  Dtype  
---  ------                            --------------  -----  
 0   PATIENT_VISIT_IDENTIFIER          351 non-null    int64  
 1   AGE_ABOVE65                       351 non-null    int64  
 2   AGE_PERCENTIL                     351 non-null    int64  
 3   GENDER                            351 non-null    int64  
 4   DISEASE GROUPING 1                351 non-null    float64
 5   DISEASE GROUPING 2                351 non-null    float64
 6   DISEASE GROUPING 3                351 non-null    float64
 7   DISEASE GROUPING 4                351 non-null    float64
 8   DISEASE GROUPING 5                351 non-null    float64
 9   DISEASE GROUPING 6                351 non-null    float64
 10  HTN                               351 non-null    float64
 11  IMMUNOCOMPROMISED                 351 non-null    float64
 12  OTHER                             351 non-null    float64
 13  ALBUMIN_MEDIAN                    351 non-null    float64
 14  BE_ARTERIAL_MEDIAN                351 non-null    float64
 15  BE_VENOUS_MEDIAN                  351 non-null    float64
 16  BIC_ARTERIAL_MEDIAN               351 non-null    float64
 17  BIC_VENOUS_MEDIAN                 351 non-null    float64
 18  BILLIRUBIN_MEDIAN                 351 non-null    float64
 19  BLAST_MEDIAN                      351 non-null    float64
 20  CALCIUM_MEDIAN                    351 non-null    float64
 21  CREATININ_MEDIAN                  351 non-null    float64
 22  FFA_MEDIAN                        351 non-null    float64
 23  GGT_MEDIAN                        351 non-null    float64
 24  GLUCOSE_MEDIAN                    351 non-null    float64
 25  HEMATOCRITE_MEDIAN                351 non-null    float64
 26  HEMOGLOBIN_MEDIAN                 351 non-null    float64
 27  INR_MEDIAN                        351 non-null    float64
 28  LACTATE_MEDIAN                    351 non-null    float64
 29  LEUKOCYTES_MEDIAN                 351 non-null    float64
 30  LINFOCITOS_MEDIAN                 351 non-null    float64
 31  NEUTROPHILES_MEDIAN               351 non-null    float64
 32  P02_ARTERIAL_MEDIAN               351 non-null    float64
 33  P02_VENOUS_MEDIAN                 351 non-null    float64
 34  PC02_ARTERIAL_MEDIAN              351 non-null    float64
 35  PC02_VENOUS_MEDIAN                351 non-null    float64
 36  PCR_MEDIAN                        351 non-null    float64
 37  PH_ARTERIAL_MEDIAN                351 non-null    float64
 38  PH_VENOUS_MEDIAN                  351 non-null    float64
 39  PLATELETS_MEDIAN                  351 non-null    float64
 40  POTASSIUM_MEDIAN                  351 non-null    float64
 41  SAT02_ARTERIAL_MEDIAN             351 non-null    float64
 42  SAT02_VENOUS_MEDIAN               351 non-null    float64
 43  SODIUM_MEDIAN                     351 non-null    float64
 44  TGO_MEDIAN                        351 non-null    float64
 45  TGP_MEDIAN                        351 non-null    float64
 46  TTPA_MEDIAN                       351 non-null    float64
 47  UREA_MEDIAN                       351 non-null    float64
 48  DIMER_MEDIAN                      351 non-null    float64
 49  BLOODPRESSURE_DIASTOLIC_MEAN      351 non-null    float64
 50  BLOODPRESSURE_SISTOLIC_MEAN       351 non-null    float64
 51  HEART_RATE_MEAN                   351 non-null    float64
 52  RESPIRATORY_RATE_MEAN             351 non-null    float64
 53  TEMPERATURE_MEAN                  351 non-null    float64
 54  OXYGEN_SATURATION_MEAN            351 non-null    float64
 55  BLOODPRESSURE_DIASTOLIC_MEDIAN    351 non-null    float64
 56  BLOODPRESSURE_SISTOLIC_MEDIAN     351 non-null    float64
 57  HEART_RATE_MEDIAN                 351 non-null    float64
 58  RESPIRATORY_RATE_MEDIAN           351 non-null    float64
 59  TEMPERATURE_MEDIAN                351 non-null    float64
 60  OXYGEN_SATURATION_MEDIAN          351 non-null    float64
 61  BLOODPRESSURE_DIASTOLIC_MIN       351 non-null    float64
 62  BLOODPRESSURE_SISTOLIC_MIN        351 non-null    float64
 63  HEART_RATE_MIN                    351 non-null    float64
 64  RESPIRATORY_RATE_MIN              351 non-null    float64
 65  TEMPERATURE_MIN                   351 non-null    float64
 66  OXYGEN_SATURATION_MIN             351 non-null    float64
 67  BLOODPRESSURE_DIASTOLIC_MAX       351 non-null    float64
 68  BLOODPRESSURE_SISTOLIC_MAX        351 non-null    float64
 69  HEART_RATE_MAX                    351 non-null    float64
 70  RESPIRATORY_RATE_MAX              351 non-null    float64
 71  TEMPERATURE_MAX                   351 non-null    float64
 72  OXYGEN_SATURATION_MAX             351 non-null    float64
 73  BLOODPRESSURE_DIASTOLIC_DIFF      351 non-null    float64
 74  BLOODPRESSURE_SISTOLIC_DIFF       351 non-null    float64
 75  HEART_RATE_DIFF                   351 non-null    float64
 76  RESPIRATORY_RATE_DIFF             351 non-null    float64
 77  TEMPERATURE_DIFF                  351 non-null    float64
 78  OXYGEN_SATURATION_DIFF            351 non-null    float64
 79  BLOODPRESSURE_DIASTOLIC_DIFF_REL  351 non-null    float64
 80  BLOODPRESSURE_SISTOLIC_DIFF_REL   351 non-null    float64
 81  HEART_RATE_DIFF_REL               351 non-null    float64
 82  RESPIRATORY_RATE_DIFF_REL         351 non-null    float64
 83  TEMPERATURE_DIFF_REL              351 non-null    float64
 84  OXYGEN_SATURATION_DIFF_REL        351 non-null    float64
 85  GO_ICU                            351 non-null    int64  
dtypes: float64(81), int64(5)
memory usage: 236.0 KB


ICU_prop_main = dataset.groupby('GO_ICU')['PATIENT_VISIT_IDENTIFIER'].count().reset_index()

labels = ["not-admitted", "admitted"]
plt.title('ICU admissions proportion', fontdict= {'fontsize' : 16}, pad=100)
plt.pie(ICU_prop_main['PATIENT_VISIT_IDENTIFIER'],textprops={'fontsize': 12},radius =2, labels = labels, startangle=90, autopct=lambda p : '{:.2f}%  ({:,.0f}patients)'.format(p,p * sum(ICU_prop_main['PATIENT_VISIT_IDENTIFIER'])/100))
plt.show()


#Age distribution according to AGE_PERCENTILE

AGE_prop_percentil = dataset.groupby('AGE_PERCENTIL')['PATIENT_VISIT_IDENTIFIER'].count().reset_index()
AGE_prop_percentil.head()
plt.figure(figsize=(9,6))
plt.bar(AGE_prop_percentil["AGE_PERCENTIL"],AGE_prop_percentil["PATIENT_VISIT_IDENTIFIER"], width = 8)
plt.xticks(AGE_prop_percentil["AGE_PERCENTIL"])
plt.ylabel("Patients Count")
plt.xlabel("Percentile")
plt.title('Age distribution according to AGE_PERCENTILE', fontdict= {'fontsize' : 14})
plt.show()


def create_bar_to_icu(feature_name):
    ax = sns.countplot(x=feature_name, hue="GO_ICU", data=dataset)
    for i in ax.containers:
        ax.bar_label(i,)
        
create_bar_to_icu("AGE_ABOVE65")


AGE_65_ICU = dataset[dataset['GO_ICU'] == 1]
AGE_65_ICU = AGE_65_ICU.groupby('AGE_ABOVE65')['PATIENT_VISIT_IDENTIFIER'].count().reset_index()

labels = ["Below-65", "Above-65"]
plt.title('ICU admissions proportion for ages below/above 65', fontdict= {'fontsize' : 16}, pad=100)
plt.pie(AGE_65_ICU['PATIENT_VISIT_IDENTIFIER'],textprops={'fontsize': 12},radius =2, labels = labels, startangle=90, autopct=lambda p : '{:.2f}%  ({:,.0f}patients)'.format(p,p * sum(AGE_65_ICU['PATIENT_VISIT_IDENTIFIER'])/100))
plt.show()


AGE_prop_percentil = dataset[dataset['GO_ICU'] == 1]
AGE_prop_percentil = AGE_prop_percentil.groupby('AGE_PERCENTIL')['PATIENT_VISIT_IDENTIFIER'].count().reset_index()
AGE_prop_percentil.head()
plt.figure(figsize=(9,6))
plt.bar(AGE_prop_percentil["AGE_PERCENTIL"],AGE_prop_percentil["PATIENT_VISIT_IDENTIFIER"], width = 8)
plt.xticks(AGE_prop_percentil["AGE_PERCENTIL"])
plt.ylabel("Patients Count")
plt.xlabel("Age Percentile")
plt.title('ICU admissions proportion according to AGE_PERCENTILE', fontdict= {'fontsize' : 14})
plt.show()


def corr_heatmap(sample, masking=False):
    sns.set_style('whitegrid')
    # Create color map ranging between two colors
    cmap = sns.diverging_palette(50, 10, as_cmap=True)
    fig, ax = plt.subplots(figsize=(10,10))
    
    if masking==False:
        correlations = sample.corr()
        sns.heatmap(correlations, cmap=cmap, vmax=1.0, center=0, fmt='.2f', 
                    square=True, linewidths=.5, annot=True, cbar_kws={"shrink": .75})
    else:
        correlations = np.triu(sample.corr())
        sns.heatmap(sample.corr(), cmap=cmap, vmax=1.0, center=0, fmt='.2f',
                square=True, linewidths=.5, annot=True, cbar_kws={"shrink": .75}, 
                    mask=correlations)
    plt.show();


sample = dataset
var = ['AGE_PERCENTIL', 'AGE_ABOVE65', 'GO_ICU']
sample = sample[var]
corr_heatmap(sample)


dataset.drop(columns = 'AGE_ABOVE65', inplace=True)


create_bar_to_icu("GENDER")


create_bar_to_icu("HTN")


create_bar_to_icu("IMMUNOCOMPROMISED")


create_bar_to_icu("OTHER")


col = ["OTHER", "IMMUNOCOMPROMISED", "HTN"]
dataset.drop(columns = col, inplace=True)


disease_groups = meta[meta.category == 'disease_group'].index
fig, axes = plt.subplots(ncols=2,nrows=3,figsize=(15,10))
col = 0
for i in range(len(disease_groups)):
    x = sns.countplot(x=disease_groups[i], hue="GO_ICU", data=dataset, ax = axes[int((i)/2)][col])
    col = (col+1) % 2


dataset.drop(columns = disease_groups, inplace=True)


vital_signs = meta[meta.category == 'vital_signs'].index
vital_signs_ICU = vital_signs.append(meta[meta.use == 'target'].index)
vital_signs_df = dataset[vital_signs_ICU]


f,ax = plt.subplots(figsize=(23,23))
sns.heatmap(vital_signs_df.drop(columns=['GO_ICU']).corr())
plt.show()


rows_list = []
for f in vital_signs:
    v1 = vital_signs_df[ vital_signs_df['GO_ICU']==1 ][f].var()
    v0 = vital_signs_df[ vital_signs_df['GO_ICU']==0 ][f].var()
    dict1 = {
        'feature' : f,
        'icu' : v1,
        'non-icu' : v0
    }
    rows_list.append(dict1)
    
var_vital_signs = pd.DataFrame(rows_list)


fig = plt.subplots(figsize=(20,8))
ax = sns.lineplot(data=var_vital_signs, x="feature", y="icu", color='r')
sns.lineplot(data=var_vital_signs, x="feature", y="non-icu", color='g')
plt.xticks(rotation=90)
plt.show()


print(f"final dataset shape is:{dataset.shape}")

final dataset shape is:(351, 76)


dataset.drop(columns = 'PATIENT_VISIT_IDENTIFIER', inplace=True)


dataset.describe()


from sklearn.feature_selection import VarianceThreshold
selector = VarianceThreshold(threshold=0.04) # Setting threshold of 4% low variance features
selector.fit(dataset.drop(['GO_ICU'], axis=1)) # Fit to train without target variables

f = np.vectorize(lambda x : not x) # Function to toggle boolean array elements

low_variance = dataset.drop(['GO_ICU'], axis=1).columns[f(selector.get_support())]
print('{} variables have too low variance.'.format(len(low_variance)))
print('These variables are {}'.format(list(low_variance)))

49 variables have too low variance.
These variables are ['ALBUMIN_MEDIAN', 'BE_ARTERIAL_MEDIAN', 'BE_VENOUS_MEDIAN', 'BIC_ARTERIAL_MEDIAN', 'BIC_VENOUS_MEDIAN', 'BILLIRUBIN_MEDIAN', 'BLAST_MEDIAN', 'CALCIUM_MEDIAN', 'CREATININ_MEDIAN', 'FFA_MEDIAN', 'GGT_MEDIAN', 'GLUCOSE_MEDIAN', 'INR_MEDIAN', 'LEUKOCYTES_MEDIAN', 'LINFOCITOS_MEDIAN', 'NEUTROPHILES_MEDIAN', 'P02_ARTERIAL_MEDIAN', 'P02_VENOUS_MEDIAN', 'PC02_ARTERIAL_MEDIAN', 'PC02_VENOUS_MEDIAN', 'PH_ARTERIAL_MEDIAN', 'PH_VENOUS_MEDIAN', 'POTASSIUM_MEDIAN', 'SAT02_ARTERIAL_MEDIAN', 'SODIUM_MEDIAN', 'TGO_MEDIAN', 'TGP_MEDIAN', 'TTPA_MEDIAN', 'UREA_MEDIAN', 'DIMER_MEDIAN', 'RESPIRATORY_RATE_MEAN', 'OXYGEN_SATURATION_MEAN', 'RESPIRATORY_RATE_MEDIAN', 'OXYGEN_SATURATION_MEDIAN', 'RESPIRATORY_RATE_MIN', 'TEMPERATURE_MIN', 'OXYGEN_SATURATION_MIN', 'BLOODPRESSURE_DIASTOLIC_MAX', 'OXYGEN_SATURATION_MAX', 'BLOODPRESSURE_DIASTOLIC_DIFF', 'BLOODPRESSURE_SISTOLIC_DIFF', 'HEART_RATE_DIFF', 'TEMPERATURE_DIFF', 'OXYGEN_SATURATION_DIFF', 'BLOODPRESSURE_DIASTOLIC_DIFF_REL', 'BLOODPRESSURE_SISTOLIC_DIFF_REL', 'HEART_RATE_DIFF_REL', 'TEMPERATURE_DIFF_REL', 'OXYGEN_SATURATION_DIFF_REL']


dataset.drop(columns = low_variance, inplace=True)


dataset.shape

(351, 26)


dataset.columns

Index(['AGE_PERCENTIL', 'GENDER', 'HEMATOCRITE_MEDIAN', 'HEMOGLOBIN_MEDIAN',
       'LACTATE_MEDIAN', 'PCR_MEDIAN', 'PLATELETS_MEDIAN',
       'SAT02_VENOUS_MEDIAN', 'BLOODPRESSURE_DIASTOLIC_MEAN',
       'BLOODPRESSURE_SISTOLIC_MEAN', 'HEART_RATE_MEAN', 'TEMPERATURE_MEAN',
       'BLOODPRESSURE_DIASTOLIC_MEDIAN', 'BLOODPRESSURE_SISTOLIC_MEDIAN',
       'HEART_RATE_MEDIAN', 'TEMPERATURE_MEDIAN',
       'BLOODPRESSURE_DIASTOLIC_MIN', 'BLOODPRESSURE_SISTOLIC_MIN',
       'HEART_RATE_MIN', 'BLOODPRESSURE_SISTOLIC_MAX', 'HEART_RATE_MAX',
       'RESPIRATORY_RATE_MAX', 'TEMPERATURE_MAX', 'RESPIRATORY_RATE_DIFF',
       'RESPIRATORY_RATE_DIFF_REL', 'GO_ICU'],
      dtype='object')


# Features most important for prediction
corr_features = dataset.corr()['GO_ICU'].sort_values(ascending=False).head(20).index[1:]


corr_features

Index(['RESPIRATORY_RATE_MAX', 'RESPIRATORY_RATE_DIFF',
       'RESPIRATORY_RATE_DIFF_REL', 'AGE_PERCENTIL', 'PCR_MEDIAN',
       'BLOODPRESSURE_SISTOLIC_MAX', 'BLOODPRESSURE_SISTOLIC_MEAN',
       'TEMPERATURE_MAX', 'BLOODPRESSURE_SISTOLIC_MEDIAN',
       'TEMPERATURE_MEDIAN', 'TEMPERATURE_MEAN', 'HEART_RATE_MAX',
       'BLOODPRESSURE_SISTOLIC_MIN', 'HEART_RATE_MEAN', 'HEART_RATE_MEDIAN',
       'HEART_RATE_MIN', 'PLATELETS_MEDIAN', 'SAT02_VENOUS_MEDIAN', 'GENDER'],
      dtype='object')


sample = dataset
var = corr_features
sample = sample[var]
corr_heatmap(sample)


corr_features

Index(['RESPIRATORY_RATE_MAX', 'RESPIRATORY_RATE_DIFF',
       'RESPIRATORY_RATE_DIFF_REL', 'AGE_PERCENTIL', 'PCR_MEDIAN',
       'BLOODPRESSURE_SISTOLIC_MAX', 'BLOODPRESSURE_SISTOLIC_MEAN',
       'TEMPERATURE_MAX', 'BLOODPRESSURE_SISTOLIC_MEDIAN',
       'TEMPERATURE_MEDIAN', 'TEMPERATURE_MEAN', 'HEART_RATE_MAX',
       'BLOODPRESSURE_SISTOLIC_MIN', 'HEART_RATE_MEAN', 'HEART_RATE_MEDIAN',
       'HEART_RATE_MIN', 'PLATELETS_MEDIAN', 'SAT02_VENOUS_MEDIAN', 'GENDER'],
      dtype='object')


X = dataset[corr_features]
y = dataset["GO_ICU"]


# HEAD(20) - 85.5% - Random Forest
# col = ['RESPIRATORY_RATE_DIFF', 'RESPIRATORY_RATE_DIFF_REL', 'BLOODPRESSURE_SISTOLIC_MEAN', 'BLOODPRESSURE_SISTOLIC_MEDIAN', "TEMPERATURE_MEAN", "HEART_RATE_MEAN", "HEART_RATE_MEDIAN", "HEART_RATE_MIN","BLOODPRESSURE_SISTOLIC_MIN"]
# X.drop(columns = col, inplace=True)


# HEAD(20) - 93.4% - Random Forest
col = ["TEMPERATURE_MEAN", "HEART_RATE_MEAN", "HEART_RATE_MEDIAN", "HEART_RATE_MIN","BLOODPRESSURE_SISTOLIC_MIN"]
X.drop(columns = col, inplace=True)

/var/folders/94/vt0_m_ds4ldfdtdt98l0j_cr0000gn/T/ipykernel_83979/2721181433.py:3: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy


sample = X
corr_heatmap(sample)


X.columns

Index(['RESPIRATORY_RATE_MAX', 'RESPIRATORY_RATE_DIFF',
       'RESPIRATORY_RATE_DIFF_REL', 'AGE_PERCENTIL', 'PCR_MEDIAN',
       'BLOODPRESSURE_SISTOLIC_MAX', 'BLOODPRESSURE_SISTOLIC_MEAN',
       'TEMPERATURE_MAX', 'BLOODPRESSURE_SISTOLIC_MEDIAN',
       'TEMPERATURE_MEDIAN', 'HEART_RATE_MAX', 'PLATELETS_MEDIAN',
       'SAT02_VENOUS_MEDIAN', 'GENDER'],
      dtype='object')


X.shape

(351, 14)


ax = sns.countplot(x="GO_ICU", data=dataset)
for i in ax.containers:
    ax.bar_label(i,)


from imblearn.over_sampling import RandomOverSampler
ros = RandomOverSampler(random_state=0)
X, y = ros.fit_resample(X, y)
ax = sns.countplot(x=y, data=dataset)
for i in ax.containers:
    ax.bar_label(i,)


norm = MinMaxScaler().fit(X)
X_norm = norm.transform(X)
X_train,X_test,y_train,y_test = train_test_split(X_norm,y,test_size=0.2,random_state=42)


clfs = {"LogisticRegression":LogisticRegression(), 
        "SVM":SVC(kernel='rbf', probability=True),
        "Decision":DecisionTreeClassifier(), 
        "RandomForest":RandomForestClassifier(), 
        "GradientBoost":GradientBoostingClassifier(),
        "XGBoost":XGBClassifier(verbosity=0), 
        "CatBoost":CatBoostClassifier(verbose=False)}


def model_fit(clfs):
    fitted_model={}
    model_result = pd.DataFrame()
    for model_name, model in clfs.items():
        model.fit(X_train,y_train)
        fitted_model.update({model_name:model})
        y_pred = model.predict(X_test)
        model_dict = {}
        model_dict['1.Algorithm'] = model_name
        model_dict['2.Accuracy'] = round(accuracy_score(y_test, y_pred),3)
        model_dict['3.Precision'] = round(precision_score(y_test, y_pred),3)
        model_dict['4.Recall'] = round(recall_score(y_test, y_pred),3)
        model_dict['5.F1'] = round(f1_score(y_test, y_pred),3)
        model_dict['6.ROC'] = round(roc_auc_score(y_test, y_pred),3)
        model_result = model_result.append(model_dict,ignore_index=True)
    return fitted_model, model_result


fitted_model, model_result = model_fit(clfs)

/var/folders/94/vt0_m_ds4ldfdtdt98l0j_cr0000gn/T/ipykernel_83979/111864206.py:15: FutureWarning:

The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.

/var/folders/94/vt0_m_ds4ldfdtdt98l0j_cr0000gn/T/ipykernel_83979/111864206.py:15: FutureWarning:

The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.

/var/folders/94/vt0_m_ds4ldfdtdt98l0j_cr0000gn/T/ipykernel_83979/111864206.py:15: FutureWarning:

The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.

/var/folders/94/vt0_m_ds4ldfdtdt98l0j_cr0000gn/T/ipykernel_83979/111864206.py:15: FutureWarning:

The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.

/var/folders/94/vt0_m_ds4ldfdtdt98l0j_cr0000gn/T/ipykernel_83979/111864206.py:15: FutureWarning:

The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.

/var/folders/94/vt0_m_ds4ldfdtdt98l0j_cr0000gn/T/ipykernel_83979/111864206.py:15: FutureWarning:

The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.

/var/folders/94/vt0_m_ds4ldfdtdt98l0j_cr0000gn/T/ipykernel_83979/111864206.py:15: FutureWarning:

The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.


model_result.sort_values(by=['2.Accuracy'],ascending=False)


model_result["1.Algorithm"][2:]

2         Decision
3     RandomForest
4    GradientBoost
5          XGBoost
6         CatBoost
Name: 1.Algorithm, dtype: object


model_name = model_result["1.Algorithm"][2:]
for model in model_name:
    feat_imp = pd.Series(fitted_model.get(model).feature_importances_, index=X.columns)
    feat_imp = feat_imp.nlargest(15).sort_values()
    feat_imp.plot(kind="barh", title="Feature Importance ({:})".format(model))
    plt.show()

	PATIENT_VISIT_IDENTIFIER	AGE_ABOVE65	AGE_PERCENTIL	GENDER	DISEASE GROUPING 5	DISEASE GROUPING 6	HTN	IMMUNOCOMPROMISED	OTHER	ALBUMIN_MEDIAN	ALBUMIN_MEAN	ALBUMIN_MIN	ALBUMIN_MAX	ALBUMIN_DIFF	BE_ARTERIAL_MEDIAN	BE_ARTERIAL_MEAN	BE_ARTERIAL_MIN	BE_ARTERIAL_MAX	BE_ARTERIAL_DIFF	BE_VENOUS_MEDIAN	BE_VENOUS_MEAN	BE_VENOUS_MIN	BE_VENOUS_MAX	BE_VENOUS_DIFF	BIC_ARTERIAL_MEDIAN	BIC_ARTERIAL_MEAN	BIC_ARTERIAL_MIN	BIC_ARTERIAL_MAX	BIC_ARTERIAL_DIFF	BIC_VENOUS_MEDIAN	BIC_VENOUS_MEAN	BIC_VENOUS_MIN	BIC_VENOUS_MAX	BIC_VENOUS_DIFF	BILLIRUBIN_MEDIAN	BILLIRUBIN_MEAN	BILLIRUBIN_MIN	BILLIRUBIN_MAX	BILLIRUBIN_DIFF	BLAST_MEDIAN	BLAST_MEAN	BLAST_MIN	BLAST_MAX	BLAST_DIFF	CALCIUM_MEDIAN	CALCIUM_MEAN	CALCIUM_MIN	CALCIUM_MAX	CALCIUM_DIFF	CREATININ_MEDIAN	CREATININ_MEAN	CREATININ_MIN	CREATININ_MAX	CREATININ_DIFF	FFA_MEDIAN	FFA_MEAN	FFA_MIN	FFA_MAX	FFA_DIFF	GGT_MEDIAN	GGT_MEAN	GGT_MIN	GGT_MAX	GGT_DIFF	GLUCOSE_MEDIAN	GLUCOSE_MEAN	GLUCOSE_MIN	GLUCOSE_MAX	GLUCOSE_DIFF	HEMATOCRITE_MEDIAN	HEMATOCRITE_MEAN	HEMATOCRITE_MIN	HEMATOCRITE_MAX	HEMATOCRITE_DIFF	HEMOGLOBIN_MEDIAN	HEMOGLOBIN_MEAN	HEMOGLOBIN_MIN	HEMOGLOBIN_MAX	HEMOGLOBIN_DIFF	INR_MEDIAN	INR_MEAN	INR_MIN	INR_MAX	INR_DIFF	LACTATE_MEDIAN	LACTATE_MEAN	LACTATE_MIN	LACTATE_MAX	LACTATE_DIFF	LEUKOCYTES_MEDIAN	LEUKOCYTES_MEAN	LEUKOCYTES_MIN	LEUKOCYTES_MAX	LEUKOCYTES_DIFF	LINFOCITOS_MEDIAN	LINFOCITOS_MEAN	LINFOCITOS_MIN	LINFOCITOS_MAX	LINFOCITOS_DIFF	NEUTROPHILES_MEDIAN	NEUTROPHILES_MEAN	NEUTROPHILES_MIN	NEUTROPHILES_MAX	NEUTROPHILES_DIFF	P02_ARTERIAL_MEDIAN	P02_ARTERIAL_MEAN	P02_ARTERIAL_MIN	P02_ARTERIAL_MAX	P02_ARTERIAL_DIFF	P02_VENOUS_MEDIAN	P02_VENOUS_MEAN	P02_VENOUS_MIN	P02_VENOUS_MAX	P02_VENOUS_DIFF	PC02_ARTERIAL_MEDIAN	PC02_ARTERIAL_MEAN	PC02_ARTERIAL_MIN	PC02_ARTERIAL_MAX	PC02_ARTERIAL_DIFF	PC02_VENOUS_MEDIAN	PC02_VENOUS_MEAN	PC02_VENOUS_MIN	PC02_VENOUS_MAX	PC02_VENOUS_DIFF	PCR_MEDIAN	PCR_MEAN	PCR_MIN	PCR_MAX	PCR_DIFF	PH_ARTERIAL_MEDIAN	PH_ARTERIAL_MEAN	PH_ARTERIAL_MIN	PH_ARTERIAL_MAX	PH_ARTERIAL_DIFF	PH_VENOUS_MEDIAN	PH_VENOUS_MEAN	PH_VENOUS_MIN	PH_VENOUS_MAX	PH_VENOUS_DIFF	PLATELETS_MEDIAN	PLATELETS_MEAN	PLATELETS_MIN	PLATELETS_MAX	PLATELETS_DIFF	POTASSIUM_MEDIAN	POTASSIUM_MEAN	POTASSIUM_MIN	POTASSIUM_MAX	POTASSIUM_DIFF	SAT02_ARTERIAL_MEDIAN	SAT02_ARTERIAL_MEAN	SAT02_ARTERIAL_MIN	SAT02_ARTERIAL_MAX	SAT02_ARTERIAL_DIFF	SAT02_VENOUS_MEDIAN	SAT02_VENOUS_MEAN	SAT02_VENOUS_MIN	SAT02_VENOUS_MAX	SAT02_VENOUS_DIFF	SODIUM_MEDIAN	SODIUM_MEAN	SODIUM_MIN	SODIUM_MAX	SODIUM_DIFF	TGO_MEDIAN	TGO_MEAN	TGO_MIN	TGO_MAX	TGO_DIFF	TGP_MEDIAN	TGP_MEAN	TGP_MIN	TGP_MAX	TGP_DIFF	TTPA_MEDIAN	TTPA_MEAN	TTPA_MIN	TTPA_MAX	TTPA_DIFF	UREA_MEDIAN	UREA_MEAN	UREA_MIN	UREA_MAX	UREA_DIFF	DIMER_MEDIAN	DIMER_MEAN	DIMER_MIN	DIMER_MAX	DIMER_DIFF	BLOODPRESSURE_DIASTOLIC_MEAN	BLOODPRESSURE_SISTOLIC_MEAN	HEART_RATE_MEAN	RESPIRATORY_RATE_MEAN	TEMPERATURE_MEAN	OXYGEN_SATURATION_MEAN	BLOODPRESSURE_DIASTOLIC_MEDIAN	BLOODPRESSURE_SISTOLIC_MEDIAN	HEART_RATE_MEDIAN	RESPIRATORY_RATE_MEDIAN	TEMPERATURE_MEDIAN	OXYGEN_SATURATION_MEDIAN	BLOODPRESSURE_DIASTOLIC_MIN	BLOODPRESSURE_SISTOLIC_MIN	HEART_RATE_MIN	RESPIRATORY_RATE_MIN	TEMPERATURE_MIN	OXYGEN_SATURATION_MIN	BLOODPRESSURE_DIASTOLIC_MAX	BLOODPRESSURE_SISTOLIC_MAX	HEART_RATE_MAX	RESPIRATORY_RATE_MAX	TEMPERATURE_MAX	OXYGEN_SATURATION_MAX	BLOODPRESSURE_DIASTOLIC_DIFF	BLOODPRESSURE_SISTOLIC_DIFF	HEART_RATE_DIFF	RESPIRATORY_RATE_DIFF	TEMPERATURE_DIFF	OXYGEN_SATURATION_DIFF	BLOODPRESSURE_DIASTOLIC_DIFF_REL	BLOODPRESSURE_SISTOLIC_DIFF_REL	HEART_RATE_DIFF_REL	RESPIRATORY_RATE_DIFF_REL	TEMPERATURE_DIFF_REL	OXYGEN_SATURATION_DIFF_REL	WINDOW	ICU	GO_ICU
0	0	1	60th	0	1.0	1.0	0.0	0.0	1.0	0.302632	0.302632	0.302632	0.302632	-1.0	-0.935829	-0.935829	-0.935829	-0.935829	-1.0	-0.931937	-0.931937	-0.931937	-0.931937	-1.0	-0.317073	-0.317073	-0.317073	-0.317073	-1.0	-0.365854	-0.365854	-0.365854	-0.365854	-1.0	-0.959009	-0.959009	-0.959009	-0.959009	-1.0	-1.0	-1.0	-1.0	-1.0	-1.0	0.255102	0.255102	0.255102	0.255102	-1.0	-0.897381	-0.897381	-0.897381	-0.897381	-1.0	-0.800640	-0.800640	-0.800640	-0.800640	-1.0	-0.807243	-0.807243	-0.807243	-0.807243	-1.0	-0.891993	-0.891993	-0.891993	-0.891993	-1.0	-0.115304	-0.115304	-0.115304	-0.115304	-1.0	-0.121951	-0.121951	-0.121951	-0.121951	-1.0	-0.956085	-0.956085	-0.956085	-0.956085	-1.0	0.018489	0.018489	0.018489	0.018489	-1.0	-0.799343	-0.799343	-0.799343	-0.799343	-1.0	-0.779046	-0.779046	-0.779046	-0.779046	-1.0	-0.868747	-0.868747	-0.868747	-0.868747	-1.0	-0.268293	-0.268293	-0.268293	-0.268293	-1.0	-0.467456	-0.467456	-0.467456	-0.467456	-1.0	-0.827586	-0.827586	-0.827586	-0.827586	-1.0	-0.785276	-0.785276	-0.785276	-0.785276	-1.0	-0.937618	-0.937618	-0.937618	-0.937618	-1.0	0.404255	0.404255	0.404255	0.404255	-1.0	0.378788	0.378788	0.378788	0.378788	-1.0	-0.506008	-0.506008	-0.506008	-0.506008	-1.0	-0.592593	-0.592593	-0.592593	-0.592593	-1.0	0.893939	0.893939	0.893939	0.893939	-1.0	0.635802	0.635802	0.635802	0.635802	-1.0	0.057143	0.057143	0.057143	0.057143	-1.0	-0.998134	-0.998134	-0.998134	-0.998134	-1.0	-0.987424	-0.987424	-0.987424	-0.987424	-1.0	-0.836123	-0.836123	-0.836123	-0.836123	-1.0	-0.836145	-0.836145	-0.836145	-0.836145	-1.0	-0.995837	-0.995837	-0.995837	-0.995837	-1.0	0.086420	-0.230769	-0.283019	-0.593220	-0.285714	0.736842	0.086420	-0.230769	-0.283019	-0.586207	-0.285714	0.736842	0.237113	0.000000	-0.162393	-0.500000	0.208791	0.898990	-0.247863	-0.459459	-0.432836	-0.636364	-0.420290	0.736842	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	0-2	0	1
5	1	1	90th	1	0.0	0.0	1.0	1.0	1.0	0.333333	0.333333	0.333333	0.333333	-1.0	-0.918004	-0.918004	-0.918004	-0.918004	-1.0	-0.972077	-0.972077	-0.972077	-0.972077	-1.0	-0.268293	-0.268293	-0.268293	-0.268293	-1.0	-0.268293	-0.268293	-0.268293	-0.268293	-1.0	-0.948485	-0.948485	-0.948485	-0.948485	-1.0	-1.0	-1.0	-1.0	-1.0	-1.0	0.476190	0.476190	0.476190	0.476190	-1.0	-0.652883	-0.652883	-0.652883	-0.652883	-1.0	-0.751244	-0.751244	-0.751244	-0.751244	-1.0	-0.966511	-0.966511	-0.966511	-0.966511	-1.0	-0.891993	-0.891993	-0.891993	-0.891993	-1.0	-0.519217	-0.519217	-0.519217	-0.519217	-1.0	-0.538064	-0.538064	-0.538064	-0.538064	-1.0	-0.973233	-0.973233	-0.973233	-0.973233	-1.0	-0.983893	-0.983893	-0.983893	-0.983893	-1.0	-0.654049	-0.654049	-0.654049	-0.654049	-1.0	-0.730290	-0.730290	-0.730290	-0.730290	-1.0	-0.715924	-0.715924	-0.715924	-0.715924	-1.0	0.052846	0.052846	0.052846	0.052846	-1.0	-0.092702	-0.092702	-0.092702	-0.092702	-1.0	-0.760920	-0.760920	-0.760920	-0.760920	-1.0	-0.664622	-0.664622	-0.664622	-0.664622	-1.0	-0.921991	-0.921991	-0.921991	-0.921991	-1.0	0.248227	0.248227	0.248227	0.248227	-1.0	0.222222	0.222222	0.222222	0.222222	-1.0	-0.388518	-0.388518	-0.388518	-0.388518	-1.0	-0.518519	-0.518519	-0.518519	-0.518519	-1.0	0.969697	0.969697	0.969697	0.969697	-1.0	0.641975	0.641975	0.641975	0.641975	-1.0	-0.168254	-0.168254	-0.168254	-0.168254	-1.0	-0.995957	-0.995957	-0.995957	-0.995957	-1.0	-0.989075	-0.989075	-0.989075	-0.989075	-1.0	-0.769041	-0.769041	-0.769041	-0.769041	-1.0	-0.633735	-0.633735	-0.633735	-0.633735	-1.0	-0.966593	-0.966593	-0.966593	-0.966593	-1.0	-0.283951	-0.046154	0.188679	0.830508	-0.107143	1.000000	-0.283951	-0.046154	0.188679	0.862069	-0.107143	1.000000	-0.072165	0.150000	0.264957	1.000000	0.318681	1.000000	-0.504274	-0.329730	-0.059701	0.636364	-0.275362	1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	0-2	1	1
10	2	0	10th	0	0.0	0.0	0.0	0.0	1.0	0.605263	0.605263	0.605263	0.605263	-1.0	-1.000000	-1.000000	-1.000000	-1.000000	-1.0	-1.000000	-1.000000	-1.000000	-1.000000	-1.0	-0.317073	-0.317073	-0.317073	-0.317073	-1.0	-0.317073	-0.317073	-0.317073	-0.317073	-1.0	-0.938950	-0.938950	-0.938950	-0.938950	-1.0	-1.0	-1.0	-1.0	-1.0	-1.0	0.357143	0.357143	0.357143	0.357143	-1.0	-0.912243	-0.912243	-0.912243	-0.912243	-1.0	-0.742004	-0.742004	-0.742004	-0.742004	-1.0	-0.958528	-0.958528	-0.958528	-0.958528	-1.0	-0.780261	-0.780261	-0.780261	-0.780261	-1.0	0.144654	0.144654	0.144654	0.144654	-1.0	0.158537	0.158537	0.158537	0.158537	-1.0	-0.959849	-0.959849	-0.959849	-0.959849	-1.0	1.000000	1.000000	1.000000	1.000000	-1.0	-0.382773	-0.382773	-0.382773	-0.382773	-1.0	-0.908714	-0.908714	-0.908714	-0.908714	-1.0	-0.412965	-0.412965	-0.412965	-0.412965	-1.0	-0.170732	-0.170732	-0.170732	-0.170732	-1.0	-0.704142	-0.704142	-0.704142	-0.704142	-1.0	-0.779310	-0.779310	-0.779310	-0.779310	-1.0	-0.754601	-0.754601	-0.754601	-0.754601	-1.0	-0.939887	-0.939887	-0.939887	-0.939887	-1.0	0.234043	0.234043	0.234043	0.234043	-1.0	0.363636	0.363636	0.363636	0.363636	-1.0	-0.399199	-0.399199	-0.399199	-0.399199	-1.0	-0.703704	-0.703704	-0.703704	-0.703704	-1.0	0.939394	0.939394	0.939394	0.939394	-1.0	0.345679	0.345679	0.345679	0.345679	-1.0	0.085714	0.085714	0.085714	0.085714	-1.0	-0.995428	-0.995428	-0.995428	-0.995428	-1.0	-0.986662	-0.986662	-0.986662	-0.986662	-1.0	-0.846633	-0.846633	-0.846633	-0.846633	-1.0	-0.836145	-0.836145	-0.836145	-0.836145	-1.0	-0.978029	-0.978029	-0.978029	-0.978029	-1.0	-0.427812	-0.688640	-0.097048	-0.636692	0.207217	0.844055	-0.423868	-0.733333	-0.100629	-0.586207	0.232143	0.850877	-0.443299	-0.541667	-0.259259	-0.666667	0.373626	0.878788	-0.418803	-0.535135	-0.049751	-0.575758	0.101449	0.982456	-0.600000	-0.554192	-0.521628	-0.803922	-0.750000	-0.885522	-0.595604	-0.419448	-0.681860	-0.792832	-0.752732	-0.887561	0-2	0	1
15	3	0	40th	1	0.0	0.0	0.0	1.0	1.0	-0.263158	-0.263158	-0.263158	-0.263158	-1.0	-1.000000	-1.000000	-1.000000	-1.000000	-1.0	-1.000000	-1.000000	-1.000000	-1.000000	-1.0	-0.317073	-0.317073	-0.317073	-0.317073	-1.0	-0.317073	-0.317073	-0.317073	-0.317073	-1.0	-0.972789	-0.972789	-0.972789	-0.972789	-1.0	-1.0	-1.0	-1.0	-1.0	-1.0	0.326531	0.326531	0.326531	0.326531	-1.0	-0.968861	-0.968861	-0.968861	-0.968861	-1.0	-0.194030	-0.194030	-0.194030	-0.194030	-1.0	-0.316589	-0.316589	-0.316589	-0.316589	-1.0	-0.891993	-0.891993	-0.891993	-0.891993	-1.0	-0.203354	-0.203354	-0.203354	-0.203354	-1.0	-0.219512	-0.219512	-0.219512	-0.219512	-1.0	-0.959849	-0.959849	-0.959849	-0.959849	-1.0	-0.828421	-0.828421	-0.828421	-0.828421	-1.0	-0.729239	-0.729239	-0.729239	-0.729239	-1.0	-0.836100	-0.836100	-0.836100	-0.836100	-1.0	-0.784714	-0.784714	-0.784714	-0.784714	-1.0	-0.170732	-0.170732	-0.170732	-0.170732	-1.0	-0.633136	-0.633136	-0.633136	-0.633136	-1.0	-0.779310	-0.779310	-0.779310	-0.779310	-1.0	-0.779141	-0.779141	-0.779141	-0.779141	-1.0	-0.503592	-0.503592	-0.503592	-0.503592	-1.0	0.234043	0.234043	0.234043	0.234043	-1.0	0.363636	0.363636	0.363636	0.363636	-1.0	-0.564753	-0.564753	-0.564753	-0.564753	-1.0	-0.777778	-0.777778	-0.777778	-0.777778	-1.0	0.939394	0.939394	0.939394	0.939394	-1.0	0.580247	0.580247	0.580247	0.580247	-1.0	0.200000	0.200000	0.200000	0.200000	-1.0	-0.989549	-0.989549	-0.989549	-0.989549	-1.0	-0.956555	-0.956555	-0.956555	-0.956555	-1.0	-0.846633	-0.846633	-0.846633	-0.846633	-1.0	-0.937349	-0.937349	-0.937349	-0.937349	-1.0	-0.978029	-0.978029	-0.978029	-0.978029	-1.0	-0.132620	-0.484650	-0.448553	-0.506215	-0.119762	0.652398	-0.102881	-0.482051	-0.459119	-0.494253	-0.142857	0.666667	-0.127148	-0.329167	-0.435897	-0.547619	0.216117	0.629630	-0.247863	-0.506306	-0.263682	-0.454545	-0.024155	0.754386	-0.692754	-0.730061	-0.582697	-0.784314	-0.682540	-0.723906	-0.769565	-0.685906	-0.689698	-0.776583	-0.682540	-0.724145	0-2	0	0
20	4	0	10th	0	0.0	0.0	0.0	0.0	1.0	0.605263	0.605263	0.605263	0.605263	-1.0	-1.000000	-1.000000	-1.000000	-1.000000	-1.0	-1.000000	-1.000000	-1.000000	-1.000000	-1.0	-0.317073	-0.317073	-0.317073	-0.317073	-1.0	-0.317073	-0.317073	-0.317073	-0.317073	-1.0	-0.935113	-0.935113	-0.935113	-0.935113	-1.0	-1.0	-1.0	-1.0	-1.0	-1.0	0.357143	0.357143	0.357143	0.357143	-1.0	-0.913659	-0.913659	-0.913659	-0.913659	-1.0	-0.829424	-0.829424	-0.829424	-0.829424	-1.0	-0.938084	-0.938084	-0.938084	-0.938084	-1.0	-0.851024	-0.851024	-0.851024	-0.851024	-1.0	0.358491	0.358491	0.358491	0.358491	-1.0	0.304878	0.304878	0.304878	0.304878	-1.0	-0.959849	-0.959849	-0.959849	-0.959849	-1.0	1.000000	1.000000	1.000000	1.000000	-1.0	-0.702202	-0.702202	-0.702202	-0.702202	-1.0	-0.641079	-0.641079	-0.641079	-0.641079	-1.0	-0.812725	-0.812725	-0.812725	-0.812725	-1.0	-0.170732	-0.170732	-0.170732	-0.170732	-1.0	-0.704142	-0.704142	-0.704142	-0.704142	-1.0	-0.779310	-0.779310	-0.779310	-0.779310	-1.0	-0.754601	-0.754601	-0.754601	-0.754601	-1.0	-0.990926	-0.990926	-0.990926	-0.990926	-1.0	0.234043	0.234043	0.234043	0.234043	-1.0	0.363636	0.363636	0.363636	0.363636	-1.0	-0.457944	-0.457944	-0.457944	-0.457944	-1.0	-0.592593	-0.592593	-0.592593	-0.592593	-1.0	0.939394	0.939394	0.939394	0.939394	-1.0	0.345679	0.345679	0.345679	0.345679	-1.0	0.142857	0.142857	0.142857	0.142857	-1.0	-0.998507	-0.998507	-0.998507	-0.998507	-1.0	-0.991235	-0.991235	-0.991235	-0.991235	-1.0	-0.846633	-0.846633	-0.846633	-0.846633	-1.0	-0.903614	-0.903614	-0.903614	-0.903614	-1.0	-1.000000	-1.000000	-1.000000	-1.000000	-1.0	0.236838	-0.101680	0.144491	-0.539451	0.189142	0.852390	0.251029	-0.102564	0.160377	-0.540230	0.226190	0.833333	0.237113	0.000000	0.048433	-0.500000	0.362637	0.925926	-0.076923	-0.293694	0.019900	-0.535354	0.072464	0.912281	-0.826087	-0.811861	-0.725191	-0.901961	-0.761905	-0.959596	-0.884058	-0.826611	-0.839287	-0.896057	-0.766042	-0.960291	0-2	0	0

	AGE_PERCENTIL	GENDER	ALBUMIN_MEDIAN	BE_ARTERIAL_MEDIAN	BE_VENOUS_MEDIAN	BIC_ARTERIAL_MEDIAN	BIC_VENOUS_MEDIAN	BILLIRUBIN_MEDIAN	BLAST_MEDIAN	CALCIUM_MEDIAN	CREATININ_MEDIAN	FFA_MEDIAN	GGT_MEDIAN	GLUCOSE_MEDIAN	HEMATOCRITE_MEDIAN	HEMOGLOBIN_MEDIAN	INR_MEDIAN	LACTATE_MEDIAN	LEUKOCYTES_MEDIAN	LINFOCITOS_MEDIAN	NEUTROPHILES_MEDIAN	P02_ARTERIAL_MEDIAN	P02_VENOUS_MEDIAN	PC02_ARTERIAL_MEDIAN	PC02_VENOUS_MEDIAN	PCR_MEDIAN	PH_ARTERIAL_MEDIAN	PH_VENOUS_MEDIAN	PLATELETS_MEDIAN	POTASSIUM_MEDIAN	SAT02_ARTERIAL_MEDIAN	SAT02_VENOUS_MEDIAN	SODIUM_MEDIAN	TGO_MEDIAN	TGP_MEDIAN	TTPA_MEDIAN	UREA_MEDIAN	DIMER_MEDIAN	BLOODPRESSURE_DIASTOLIC_MEAN	BLOODPRESSURE_SISTOLIC_MEAN	HEART_RATE_MEAN	RESPIRATORY_RATE_MEAN	TEMPERATURE_MEAN	OXYGEN_SATURATION_MEAN	BLOODPRESSURE_DIASTOLIC_MEDIAN	BLOODPRESSURE_SISTOLIC_MEDIAN	HEART_RATE_MEDIAN	RESPIRATORY_RATE_MEDIAN	TEMPERATURE_MEDIAN	OXYGEN_SATURATION_MEDIAN	BLOODPRESSURE_DIASTOLIC_MIN	BLOODPRESSURE_SISTOLIC_MIN	HEART_RATE_MIN	RESPIRATORY_RATE_MIN	TEMPERATURE_MIN	OXYGEN_SATURATION_MIN	BLOODPRESSURE_DIASTOLIC_MAX	BLOODPRESSURE_SISTOLIC_MAX	HEART_RATE_MAX	RESPIRATORY_RATE_MAX	TEMPERATURE_MAX	OXYGEN_SATURATION_MAX	BLOODPRESSURE_DIASTOLIC_DIFF	BLOODPRESSURE_SISTOLIC_DIFF	HEART_RATE_DIFF	RESPIRATORY_RATE_DIFF	TEMPERATURE_DIFF	OXYGEN_SATURATION_DIFF	BLOODPRESSURE_DIASTOLIC_DIFF_REL	BLOODPRESSURE_SISTOLIC_DIFF_REL	HEART_RATE_DIFF_REL	RESPIRATORY_RATE_DIFF_REL	TEMPERATURE_DIFF_REL	OXYGEN_SATURATION_DIFF_REL	GO_ICU
count	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000	351.000000
mean	52.905983	0.381766	0.554956	-0.985557	-0.949424	-0.316517	-0.317374	-0.944460	-0.993483	0.332771	-0.893073	-0.726528	-0.930316	-0.856552	-0.133667	-0.172548	-0.940769	0.449888	-0.760812	-0.736898	-0.825891	-0.177290	-0.686570	-0.778092	-0.759099	-0.834459	0.232679	0.373277	-0.474968	-0.545426	0.921804	0.313619	-0.072401	-0.993738	-0.985877	-0.830554	-0.840145	-0.952026	-0.062471	-0.327484	-0.249250	-0.463956	0.085013	0.739482	-0.065946	-0.332360	-0.250379	-0.458829	0.081629	0.743359	0.012693	-0.170462	-0.227380	-0.475258	0.356068	0.834863	-0.244617	-0.430086	-0.298933	-0.386558	0.006104	0.802169	-0.807403	-0.799338	-0.804988	-0.777917	-0.809298	-0.910798	-0.837553	-0.789700	-0.857572	-0.787122	-0.810324	-0.910752	0.461538
std	28.816458	0.486513	0.149646	0.072823	0.118326	0.055182	0.096368	0.060816	0.106031	0.078625	0.089030	0.147544	0.131480	0.104911	0.209338	0.222004	0.051919	0.734231	0.114210	0.139272	0.111417	0.090416	0.127270	0.049589	0.075473	0.200942	0.066200	0.098439	0.201132	0.151616	0.113045	0.276295	0.187040	0.017217	0.015785	0.061872	0.104319	0.132466	0.222431	0.258340	0.246603	0.178930	0.213823	0.145499	0.225319	0.258794	0.255383	0.182155	0.216619	0.147381	0.208583	0.220084	0.233925	0.188489	0.149017	0.158410	0.185468	0.208580	0.222574	0.251018	0.200912	0.149406	0.185209	0.194405	0.175738	0.236197	0.153417	0.157520	0.164183	0.199713	0.121888	0.213955	0.151840	0.157670	0.499230
min	10.000000	0.000000	-0.263158	-1.000000	-1.000000	-0.756098	-1.000000	-0.992674	-1.000000	0.030612	-0.970276	-0.927505	-0.997664	-0.929236	-0.903564	-0.871951	-0.991217	-0.975884	-0.966010	-0.977178	-0.990796	-0.878049	-0.988166	-0.958621	-1.000000	-1.000000	-0.489362	-0.318182	-0.991989	-0.962963	-0.575758	-0.925926	-0.714286	-0.999627	-0.999619	-0.961853	-0.978313	-1.000000	-0.654321	-0.880959	-0.811321	-0.932203	-0.464286	-1.000000	-0.679012	-0.884615	-0.811321	-0.931034	-0.464286	-1.000000	-0.587629	-0.862500	-0.803419	-0.964286	0.032967	-0.060606	-0.760684	-0.870270	-0.850746	-0.939394	-0.565217	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	0.000000
25%	30.000000	0.000000	0.605263	-1.000000	-1.000000	-0.317073	-0.317073	-0.962498	-1.000000	0.318027	-0.928521	-0.757996	-0.958528	-0.891993	-0.250524	-0.299797	-0.959849	0.027331	-0.834299	-0.830221	-0.897359	-0.170732	-0.704142	-0.779310	-0.760736	-0.967013	0.234043	0.363636	-0.603471	-0.629630	0.939394	0.345679	-0.171429	-0.996711	-0.992759	-0.846633	-0.896386	-0.983001	-0.192842	-0.501207	-0.407943	-0.559322	-0.077267	0.689493	-0.197531	-0.504167	-0.415094	-0.551724	-0.075893	0.688596	-0.120275	-0.314583	-0.367521	-0.571429	0.252747	0.838384	-0.361823	-0.567568	-0.445274	-0.555556	-0.130435	0.736842	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	0.000000
50%	50.000000	0.000000	0.605263	-1.000000	-1.000000	-0.317073	-0.317073	-0.941218	-1.000000	0.357143	-0.909413	-0.742004	-0.958528	-0.891993	-0.132075	-0.170732	-0.959849	1.000000	-0.788142	-0.765560	-0.851140	-0.170732	-0.704142	-0.779310	-0.754601	-0.894140	0.234043	0.363636	-0.484646	-0.555556	0.939394	0.345679	-0.028571	-0.995521	-0.988377	-0.846633	-0.869880	-0.978029	-0.061728	-0.369231	-0.266799	-0.506053	0.063492	0.747829	-0.061728	-0.369231	-0.273585	-0.505747	0.059524	0.754386	0.024055	-0.187500	-0.253561	-0.500000	0.333333	0.878788	-0.247863	-0.452252	-0.318408	-0.484848	0.007246	0.828947	-0.826087	-0.815951	-0.816794	-0.852941	-0.809524	-0.949495	-0.857708	-0.807261	-0.859410	-0.847670	-0.808466	-0.950579	0.000000
75%	80.000000	1.000000	0.605263	-1.000000	-0.950262	-0.317073	-0.317073	-0.938950	-1.000000	0.357143	-0.881104	-0.742004	-0.954634	-0.855680	-0.002096	-0.036585	-0.933501	1.000000	-0.714465	-0.676349	-0.784914	-0.170732	-0.704142	-0.779310	-0.754601	-0.784121	0.234043	0.378788	-0.375834	-0.481481	0.939394	0.345679	0.057143	-0.994448	-0.985137	-0.830673	-0.821687	-0.964518	0.077444	-0.184035	-0.128039	-0.413506	0.214675	0.812427	0.086420	-0.193590	-0.135220	-0.419540	0.214286	0.815789	0.134021	-0.062500	-0.136752	-0.428571	0.428571	0.909091	-0.128205	-0.300450	-0.171642	-0.237374	0.140097	0.877193	-0.721739	-0.705521	-0.716285	-0.647059	-0.706349	-0.909091	-0.773592	-0.687866	-0.791729	-0.664875	-0.708330	-0.909404	1.000000
max	100.000000	1.000000	0.789474	0.000000	0.000000	0.341463	0.341463	0.030525	0.969834	0.693878	0.232838	0.680171	1.000000	0.452514	0.811321	0.804878	-0.355082	1.000000	-0.020471	0.080913	-0.070028	0.731707	0.372781	0.103448	0.122699	0.535350	0.617021	0.818182	0.345794	0.222222	1.000000	1.000000	1.000000	-0.693944	-0.794970	-0.341573	0.050602	1.000000	1.000000	1.000000	0.905660	0.830508	0.964286	1.000000	1.000000	1.000000	0.962264	0.862069	0.964286	1.000000	1.000000	1.000000	0.914530	1.000000	0.978022	1.000000	0.384615	0.405405	0.552239	0.636364	0.681159	1.000000	0.130435	0.411043	0.160305	0.058824	-0.095238	-0.010101	0.130435	0.378657	-0.343264	-0.054337	-0.124649	-0.000219	1.000000

	1.Algorithm	2.Accuracy	3.Precision	4.Recall	5.F1	6.ROC
3	RandomForest	0.934	0.921	0.946	0.933	0.935
6	CatBoost	0.921	0.919	0.919	0.919	0.921
4	GradientBoost	0.908	0.895	0.919	0.907	0.908
5	XGBoost	0.908	0.917	0.892	0.904	0.907
2	Decision	0.882	0.868	0.892	0.880	0.882
0	LogisticRegression	0.816	0.829	0.784	0.806	0.815
1	SVM	0.803	0.867	0.703	0.776	0.800

	PATIENT_VISIT_IDENTIFIER	AGE_PERCENTIL	GENDER	DISEASE GROUPING 3	OTHER	ALBUMIN_MEDIAN	ALBUMIN_MEAN	ALBUMIN_MIN	ALBUMIN_MAX	ALBUMIN_DIFF	BE_ARTERIAL_MEDIAN	BE_ARTERIAL_MEAN	BE_ARTERIAL_MIN	BE_ARTERIAL_MAX	BE_ARTERIAL_DIFF	BE_VENOUS_MEDIAN	BE_VENOUS_MEAN	BE_VENOUS_MIN	BE_VENOUS_MAX	BE_VENOUS_DIFF	BIC_ARTERIAL_MEDIAN	BIC_ARTERIAL_MEAN	BIC_ARTERIAL_MIN	BIC_ARTERIAL_MAX	BIC_ARTERIAL_DIFF	BIC_VENOUS_MEDIAN	BIC_VENOUS_MEAN	BIC_VENOUS_MIN	BIC_VENOUS_MAX	BIC_VENOUS_DIFF	BILLIRUBIN_MEDIAN	BILLIRUBIN_MEAN	BILLIRUBIN_MIN	BILLIRUBIN_MAX	BILLIRUBIN_DIFF	BLAST_MEDIAN	BLAST_MEAN	BLAST_MIN	BLAST_MAX	BLAST_DIFF	CALCIUM_MEDIAN	CALCIUM_MEAN	CALCIUM_MIN	CALCIUM_MAX	CALCIUM_DIFF	CREATININ_MEDIAN	CREATININ_MEAN	CREATININ_MIN	CREATININ_MAX	CREATININ_DIFF	FFA_MEDIAN	FFA_MEAN	FFA_MIN	FFA_MAX	FFA_DIFF	GGT_MEDIAN	GGT_MEAN	GGT_MIN	GGT_MAX	GGT_DIFF	GLUCOSE_MEDIAN	GLUCOSE_MEAN	GLUCOSE_MIN	GLUCOSE_MAX	GLUCOSE_DIFF	HEMATOCRITE_MEDIAN	HEMATOCRITE_MEAN	HEMATOCRITE_MIN	HEMATOCRITE_MAX	HEMATOCRITE_DIFF	HEMOGLOBIN_MEDIAN	HEMOGLOBIN_MEAN	HEMOGLOBIN_MIN	HEMOGLOBIN_MAX	HEMOGLOBIN_DIFF	INR_MEDIAN	INR_MEAN	INR_MIN	INR_MAX	INR_DIFF	LACTATE_MEDIAN	LACTATE_MEAN	LACTATE_MIN	LACTATE_MAX	LACTATE_DIFF	LEUKOCYTES_MEDIAN	LEUKOCYTES_MEAN	LEUKOCYTES_MIN	LEUKOCYTES_MAX	LEUKOCYTES_DIFF	LINFOCITOS_MEDIAN	LINFOCITOS_MEAN	LINFOCITOS_MIN	LINFOCITOS_MAX	LINFOCITOS_DIFF	NEUTROPHILES_MEDIAN	NEUTROPHILES_MEAN	NEUTROPHILES_MIN	NEUTROPHILES_MAX	NEUTROPHILES_DIFF	P02_ARTERIAL_MEDIAN	P02_ARTERIAL_MEAN	P02_ARTERIAL_MIN	P02_ARTERIAL_MAX	P02_ARTERIAL_DIFF	P02_VENOUS_MEDIAN	P02_VENOUS_MEAN	P02_VENOUS_MIN	P02_VENOUS_MAX	P02_VENOUS_DIFF	PC02_ARTERIAL_MEDIAN	PC02_ARTERIAL_MEAN	PC02_ARTERIAL_MIN	PC02_ARTERIAL_MAX	PC02_ARTERIAL_DIFF	PC02_VENOUS_MEDIAN	PC02_VENOUS_MEAN	PC02_VENOUS_MIN	PC02_VENOUS_MAX	PC02_VENOUS_DIFF	PCR_MEDIAN	PCR_MEAN	PCR_MIN	PCR_MAX	PCR_DIFF	PH_ARTERIAL_MEDIAN	PH_ARTERIAL_MEAN	PH_ARTERIAL_MIN	PH_ARTERIAL_MAX	PH_ARTERIAL_DIFF	PH_VENOUS_MEDIAN	PH_VENOUS_MEAN	PH_VENOUS_MIN	PH_VENOUS_MAX	PH_VENOUS_DIFF	PLATELETS_MEDIAN	PLATELETS_MEAN	PLATELETS_MIN	PLATELETS_MAX	PLATELETS_DIFF	POTASSIUM_MEDIAN	POTASSIUM_MEAN	POTASSIUM_MIN	POTASSIUM_MAX	POTASSIUM_DIFF	SAT02_ARTERIAL_MEDIAN	SAT02_ARTERIAL_MEAN	SAT02_ARTERIAL_MIN	SAT02_ARTERIAL_MAX	SAT02_ARTERIAL_DIFF	SAT02_VENOUS_MEDIAN	SAT02_VENOUS_MEAN	SAT02_VENOUS_MIN	SAT02_VENOUS_MAX	SAT02_VENOUS_DIFF	SODIUM_MEDIAN	SODIUM_MEAN	SODIUM_MIN	SODIUM_MAX	SODIUM_DIFF	TGO_MEDIAN	TGO_MEAN	TGO_MIN	TGO_MAX	TGO_DIFF	TGP_MEDIAN	TGP_MEAN	TGP_MIN	TGP_MAX	TGP_DIFF	TTPA_MEDIAN	TTPA_MEAN	TTPA_MIN	TTPA_MAX	TTPA_DIFF	UREA_MEDIAN	UREA_MEAN	UREA_MIN	UREA_MAX	UREA_DIFF	DIMER_MEDIAN	DIMER_MEAN	DIMER_MIN	DIMER_MAX	DIMER_DIFF	BLOODPRESSURE_DIASTOLIC_MEAN	BLOODPRESSURE_SISTOLIC_MEAN	HEART_RATE_MEAN	RESPIRATORY_RATE_MEAN	TEMPERATURE_MEAN	OXYGEN_SATURATION_MEAN	BLOODPRESSURE_DIASTOLIC_MEDIAN	BLOODPRESSURE_SISTOLIC_MEDIAN	HEART_RATE_MEDIAN	RESPIRATORY_RATE_MEDIAN	TEMPERATURE_MEDIAN	OXYGEN_SATURATION_MEDIAN	BLOODPRESSURE_DIASTOLIC_MIN	BLOODPRESSURE_SISTOLIC_MIN	HEART_RATE_MIN	RESPIRATORY_RATE_MIN	TEMPERATURE_MIN	OXYGEN_SATURATION_MIN	BLOODPRESSURE_DIASTOLIC_MAX	BLOODPRESSURE_SISTOLIC_MAX	HEART_RATE_MAX	RESPIRATORY_RATE_MAX	TEMPERATURE_MAX	OXYGEN_SATURATION_MAX	BLOODPRESSURE_DIASTOLIC_DIFF	BLOODPRESSURE_SISTOLIC_DIFF	HEART_RATE_DIFF	RESPIRATORY_RATE_DIFF	TEMPERATURE_DIFF	OXYGEN_SATURATION_DIFF	BLOODPRESSURE_DIASTOLIC_DIFF_REL	BLOODPRESSURE_SISTOLIC_DIFF_REL	HEART_RATE_DIFF_REL	RESPIRATORY_RATE_DIFF_REL	TEMPERATURE_DIFF_REL	OXYGEN_SATURATION_DIFF_REL	WINDOW
1920	384	50th	1	0.0	1.0	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	0.012346	-0.292308	0.056604	-0.525424	0.535714	0.789474	0.012346	-0.292308	0.056604	-0.517241	0.535714	0.789474	0.175258	-0.050	0.145299	-0.428571	0.714286	0.919192	-0.299145	-0.502703	-0.164179	-0.575758	0.246377	0.789474	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	0-2
1921	384	50th	1	0.0	1.0	0.605263	0.605263	0.605263	0.605263	-1.0	-1.0	-1.0	-1.0	-1.0	-1.0	-0.717277	-0.717277	-0.717277	-0.717277	-1.0	-0.317073	-0.317073	-0.317073	-0.317073	-1.0	-0.170732	-0.170732	-0.170732	-0.170732	-1.0	-0.982208	-0.982208	-0.982208	-0.982208	-1.0	-1.0	-1.0	-1.0	-1.0	-1.0	0.244898	0.244898	0.244898	0.244898	-1.0	-0.934890	-0.934890	-0.934890	-0.934890	-1.0	-0.782516	-0.782516	-0.782516	-0.782516	-1.0	-0.960280	-0.960280	-0.960280	-0.960280	-1.0	-0.862197	-0.862197	-0.862197	-0.862197	-1.0	-0.064990	-0.064990	-0.064990	-0.064990	-1.0	-0.158537	-0.158537	-0.158537	-0.158537	-1.0	-0.957340	-0.957340	-0.957340	-0.957340	-1.0	-0.897773	-0.897773	-0.897773	-0.897773	-1.0	-0.848590	-0.848590	-0.848590	-0.848590	-1.0	-0.686722	-0.686722	-0.686722	-0.686722	-1.0	-0.913165	-0.913165	-0.913165	-0.913165	-1.0	-0.170732	-0.170732	-0.170732	-0.170732	-1.0	-0.857988	-0.857988	-0.857988	-0.857988	-1.0	-0.77931	-0.77931	-0.77931	-0.77931	-1.0	-0.730061	-0.730061	-0.730061	-0.730061	-1.0	-0.906238	-0.906238	-0.906238	-0.906238	-1.0	0.234043	0.234043	0.234043	0.234043	-1.0	0.424242	0.424242	0.424242	0.424242	-1.0	-0.479306	-0.479306	-0.479306	-0.479306	-1.0	-0.333333	-0.333333	-0.333333	-0.333333	-1.0	0.939394	0.939394	0.939394	0.939394	-1.0	-0.333333	-0.333333	-0.333333	-0.333333	-1.0	-0.085714	-0.085714	-0.085714	-0.085714	-1.0	-0.997387	-0.997387	-0.997387	-0.997387	-1.0	-0.992378	-0.992378	-0.992378	-0.992378	-1.0	-0.869210	-0.869210	-0.869210	-0.869210	-1.0	-0.879518	-0.879518	-0.879518	-0.879518	-1.0	-0.979571	-0.979571	-0.979571	-0.979571	-1.0	0.086420	-0.384615	-0.113208	-0.593220	0.142857	0.578947	0.086420	-0.384615	-0.113208	-0.586207	0.142857	0.578947	0.237113	-0.125	-0.008547	-0.500000	0.472527	0.838384	-0.247863	-0.567568	-0.298507	-0.636364	-0.072464	0.578947	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	2-4
1922	384	50th	1	0.0	1.0	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	0.086420	-0.230769	-0.169811	-0.593220	0.142857	0.736842	0.086420	-0.230769	-0.169811	-0.586207	0.142857	0.736842	0.237113	0.000	-0.059829	-0.500000	0.472527	0.898990	-0.247863	-0.459459	-0.343284	-0.636364	-0.072464	0.736842	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	4-6
1923	384	50th	1	0.0	1.0	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	0.209877	-0.384615	-0.188679	-0.661017	0.285714	0.473684	0.209877	-0.384615	-0.188679	-0.655172	0.285714	0.473684	0.340206	-0.125	-0.076923	-0.571429	0.560440	0.797980	-0.162393	-0.567568	-0.358209	-0.696970	0.043478	0.473684	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	-1.000000	6-12
1924	384	50th	1	1.0	1.0	0.605263	0.605263	0.605263	0.605263	-1.0	-1.0	-1.0	-1.0	-1.0	-1.0	-1.000000	-1.000000	-1.000000	-1.000000	-1.0	-0.317073	-0.317073	-0.317073	-0.317073	-1.0	-0.317073	-0.317073	-0.317073	-0.317073	-1.0	-0.983255	-0.983255	-0.983255	-0.983255	-1.0	-1.0	-1.0	-1.0	-1.0	-1.0	0.306122	0.306122	0.306122	0.306122	-1.0	-0.944798	-0.944798	-0.944798	-0.944798	-1.0	-0.825160	-0.825160	-0.825160	-0.825160	-1.0	-0.962617	-0.962617	-0.962617	-0.962617	-1.0	-0.891993	-0.891993	-0.891993	-0.891993	-1.0	-0.157233	-0.157233	-0.157233	-0.157233	-1.0	-0.292683	-0.292683	-0.292683	-0.292683	-1.0	-0.959849	-0.959849	-0.959849	-0.959849	-1.0	1.000000	1.000000	1.000000	1.000000	-1.0	-0.850521	-0.850521	-0.850521	-0.850521	-1.0	-0.634855	-0.634855	-0.634855	-0.634855	-1.0	-0.935974	-0.935974	-0.935974	-0.935974	-1.0	-0.170732	-0.170732	-0.170732	-0.170732	-1.0	-0.704142	-0.704142	-0.704142	-0.704142	-1.0	-0.77931	-0.77931	-0.77931	-0.77931	-1.0	-0.754601	-0.754601	-0.754601	-0.754601	-1.0	-0.801134	-0.801134	-0.801134	-0.801134	-1.0	0.234043	0.234043	0.234043	0.234043	-1.0	0.363636	0.363636	0.363636	0.363636	-1.0	-0.463284	-0.463284	-0.463284	-0.463284	-1.0	-0.444444	-0.444444	-0.444444	-0.444444	-1.0	0.939394	0.939394	0.939394	0.939394	-1.0	0.345679	0.345679	0.345679	0.345679	-1.0	0.200000	0.200000	0.200000	0.200000	-1.0	-0.997761	-0.997761	-0.997761	-0.997761	-1.0	-0.991997	-0.991997	-0.991997	-0.991997	-1.0	-0.846633	-0.846633	-0.846633	-0.846633	-1.0	-0.807229	-0.807229	-0.807229	-0.807229	-1.0	-0.888448	-0.888448	-0.888448	-0.888448	-1.0	-0.185185	-0.539103	-0.107704	-0.610169	0.050595	0.662281	-0.160494	-0.538462	-0.075472	-0.586207	0.071429	0.631579	-0.175258	-0.375	-0.247863	-0.785714	0.186813	0.777778	-0.247863	-0.470270	-0.149254	-0.515152	0.101449	0.842105	-0.652174	-0.644172	-0.633588	-0.647059	-0.547619	-0.838384	-0.701863	-0.585967	-0.763868	-0.612903	-0.551337	-0.835052	ABOVE_12

	PATIENT_VISIT_IDENTIFIER	AGE_ABOVE65	AGE_PERCENTIL	GENDER	DISEASE GROUPING 1	DISEASE GROUPING 2	DISEASE GROUPING 3	DISEASE GROUPING 4	DISEASE GROUPING 5	DISEASE GROUPING 6	HTN	IMMUNOCOMPROMISED	OTHER	ALBUMIN_MEDIAN	ALBUMIN_MEAN	ALBUMIN_MIN	ALBUMIN_MAX	ALBUMIN_DIFF	BE_ARTERIAL_MEDIAN	BE_ARTERIAL_MEAN	BE_ARTERIAL_MIN	BE_ARTERIAL_MAX	BE_ARTERIAL_DIFF	BE_VENOUS_MEDIAN	BE_VENOUS_MEAN	BE_VENOUS_MIN	BE_VENOUS_MAX	BE_VENOUS_DIFF	BIC_ARTERIAL_MEDIAN	BIC_ARTERIAL_MEAN	BIC_ARTERIAL_MIN	BIC_ARTERIAL_MAX	BIC_ARTERIAL_DIFF	BIC_VENOUS_MEDIAN	BIC_VENOUS_MEAN	BIC_VENOUS_MIN	BIC_VENOUS_MAX	BIC_VENOUS_DIFF	BILLIRUBIN_MEDIAN	BILLIRUBIN_MEAN	BILLIRUBIN_MIN	BILLIRUBIN_MAX	BILLIRUBIN_DIFF	BLAST_MEDIAN	BLAST_MEAN	BLAST_MIN	BLAST_MAX	BLAST_DIFF	CALCIUM_MEDIAN	CALCIUM_MEAN	CALCIUM_MIN	CALCIUM_MAX	CALCIUM_DIFF	CREATININ_MEDIAN	CREATININ_MEAN	CREATININ_MIN	CREATININ_MAX	CREATININ_DIFF	FFA_MEDIAN	FFA_MEAN	FFA_MIN	FFA_MAX	FFA_DIFF	GGT_MEDIAN	GGT_MEAN	GGT_MIN	GGT_MAX	GGT_DIFF	GLUCOSE_MEDIAN	GLUCOSE_MEAN	GLUCOSE_MIN	GLUCOSE_MAX	GLUCOSE_DIFF	HEMATOCRITE_MEDIAN	HEMATOCRITE_MEAN	HEMATOCRITE_MIN	HEMATOCRITE_MAX	HEMATOCRITE_DIFF	HEMOGLOBIN_MEDIAN	HEMOGLOBIN_MEAN	HEMOGLOBIN_MIN	HEMOGLOBIN_MAX	HEMOGLOBIN_DIFF	INR_MEDIAN	INR_MEAN	INR_MIN	INR_MAX	INR_DIFF	LACTATE_MEDIAN	LACTATE_MEAN	LACTATE_MIN	LACTATE_MAX	LACTATE_DIFF	LEUKOCYTES_MEDIAN	LEUKOCYTES_MEAN	LEUKOCYTES_MIN	LEUKOCYTES_MAX	LEUKOCYTES_DIFF	LINFOCITOS_MEDIAN	LINFOCITOS_MEAN	LINFOCITOS_MIN	LINFOCITOS_MAX	LINFOCITOS_DIFF	NEUTROPHILES_MEDIAN	NEUTROPHILES_MEAN	NEUTROPHILES_MIN	NEUTROPHILES_MAX	NEUTROPHILES_DIFF	P02_ARTERIAL_MEDIAN	P02_ARTERIAL_MEAN	P02_ARTERIAL_MIN	P02_ARTERIAL_MAX	P02_ARTERIAL_DIFF	P02_VENOUS_MEDIAN	P02_VENOUS_MEAN	P02_VENOUS_MIN	P02_VENOUS_MAX	P02_VENOUS_DIFF	PC02_ARTERIAL_MEDIAN	PC02_ARTERIAL_MEAN	PC02_ARTERIAL_MIN	PC02_ARTERIAL_MAX	PC02_ARTERIAL_DIFF	PC02_VENOUS_MEDIAN	PC02_VENOUS_MEAN	PC02_VENOUS_MIN	PC02_VENOUS_MAX	PC02_VENOUS_DIFF	PCR_MEDIAN	PCR_MEAN	PCR_MIN	PCR_MAX	PCR_DIFF	PH_ARTERIAL_MEDIAN	PH_ARTERIAL_MEAN	PH_ARTERIAL_MIN	PH_ARTERIAL_MAX	PH_ARTERIAL_DIFF	PH_VENOUS_MEDIAN	PH_VENOUS_MEAN	PH_VENOUS_MIN	PH_VENOUS_MAX	PH_VENOUS_DIFF	PLATELETS_MEDIAN	PLATELETS_MEAN	PLATELETS_MIN	PLATELETS_MAX	PLATELETS_DIFF	POTASSIUM_MEDIAN	POTASSIUM_MEAN	POTASSIUM_MIN	POTASSIUM_MAX	POTASSIUM_DIFF	SAT02_ARTERIAL_MEDIAN	SAT02_ARTERIAL_MEAN	SAT02_ARTERIAL_MIN	SAT02_ARTERIAL_MAX	SAT02_ARTERIAL_DIFF	SAT02_VENOUS_MEDIAN	SAT02_VENOUS_MEAN	SAT02_VENOUS_MIN	SAT02_VENOUS_MAX	SAT02_VENOUS_DIFF	SODIUM_MEDIAN	SODIUM_MEAN	SODIUM_MIN	SODIUM_MAX	SODIUM_DIFF	TGO_MEDIAN	TGO_MEAN	TGO_MIN	TGO_MAX	TGO_DIFF	TGP_MEDIAN	TGP_MEAN	TGP_MIN	TGP_MAX	TGP_DIFF	TTPA_MEDIAN	TTPA_MEAN	TTPA_MIN	TTPA_MAX	TTPA_DIFF	UREA_MEDIAN	UREA_MEAN	UREA_MIN	UREA_MAX	UREA_DIFF	DIMER_MEDIAN	DIMER_MEAN	DIMER_MIN	DIMER_MAX	DIMER_DIFF	BLOODPRESSURE_DIASTOLIC_MEAN	BLOODPRESSURE_SISTOLIC_MEAN	HEART_RATE_MEAN	RESPIRATORY_RATE_MEAN	TEMPERATURE_MEAN	OXYGEN_SATURATION_MEAN	BLOODPRESSURE_DIASTOLIC_MEDIAN	BLOODPRESSURE_SISTOLIC_MEDIAN	HEART_RATE_MEDIAN	RESPIRATORY_RATE_MEDIAN	TEMPERATURE_MEDIAN	OXYGEN_SATURATION_MEDIAN	BLOODPRESSURE_DIASTOLIC_MIN	BLOODPRESSURE_SISTOLIC_MIN	HEART_RATE_MIN	RESPIRATORY_RATE_MIN	TEMPERATURE_MIN	OXYGEN_SATURATION_MIN	BLOODPRESSURE_DIASTOLIC_MAX	BLOODPRESSURE_SISTOLIC_MAX	HEART_RATE_MAX	RESPIRATORY_RATE_MAX	TEMPERATURE_MAX	OXYGEN_SATURATION_MAX	BLOODPRESSURE_DIASTOLIC_DIFF	BLOODPRESSURE_SISTOLIC_DIFF	HEART_RATE_DIFF	RESPIRATORY_RATE_DIFF	TEMPERATURE_DIFF	OXYGEN_SATURATION_DIFF	BLOODPRESSURE_DIASTOLIC_DIFF_REL	BLOODPRESSURE_SISTOLIC_DIFF_REL	HEART_RATE_DIFF_REL	RESPIRATORY_RATE_DIFF_REL	TEMPERATURE_DIFF_REL	OXYGEN_SATURATION_DIFF_REL	GO_ICU
183	199	0	10th	0	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	0
262	287	0	10th	0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	1.0	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	-0.111111	-0.630769	0.509434	-0.254237	0.107143	0.842105	-0.111111	-0.630769	0.509434	-0.241379	0.107143	0.842105	0.072165	-0.325	0.555556	-0.142857	0.450549	0.939394	-0.384615	-0.740541	0.19403	-0.333333	-0.101449	0.842105	-1.0	-1.0	-1.0	-1.0	-1.0	-1.0	-1.0	-1.0	-1.0	-1.0	-1.0	-1.0	1

	use	type	dtype	category
varname
PATIENT_VISIT_IDENTIFIER	patient id	categorical	int64	none
AGE_ABOVE65	input	binary	int64	none
AGE_PERCENTIL	input	categorical	int64	none
GENDER	input	binary	int64	none
DISEASE GROUPING 1	input	binary	float64	disease_group
DISEASE GROUPING 2	input	binary	float64	disease_group
DISEASE GROUPING 3	input	binary	float64	disease_group
DISEASE GROUPING 4	input	binary	float64	disease_group
DISEASE GROUPING 5	input	binary	float64	disease_group
DISEASE GROUPING 6	input	binary	float64	disease_group
HTN	input	real	float64	none
IMMUNOCOMPROMISED	input	real	float64	none
OTHER	input	real	float64	none
ALBUMIN_MEDIAN	input	real	float64	vital_signs
BE_ARTERIAL_MEDIAN	input	real	float64	vital_signs
BE_VENOUS_MEDIAN	input	real	float64	vital_signs
BIC_ARTERIAL_MEDIAN	input	real	float64	vital_signs
BIC_VENOUS_MEDIAN	input	real	float64	vital_signs
BILLIRUBIN_MEDIAN	input	real	float64	vital_signs
BLAST_MEDIAN	input	real	float64	vital_signs
CALCIUM_MEDIAN	input	real	float64	vital_signs
CREATININ_MEDIAN	input	real	float64	vital_signs
FFA_MEDIAN	input	real	float64	vital_signs
GGT_MEDIAN	input	real	float64	vital_signs
GLUCOSE_MEDIAN	input	real	float64	vital_signs
HEMATOCRITE_MEDIAN	input	real	float64	vital_signs
HEMOGLOBIN_MEDIAN	input	real	float64	vital_signs
INR_MEDIAN	input	real	float64	vital_signs
LACTATE_MEDIAN	input	real	float64	vital_signs
LEUKOCYTES_MEDIAN	input	real	float64	vital_signs
LINFOCITOS_MEDIAN	input	real	float64	vital_signs
NEUTROPHILES_MEDIAN	input	real	float64	vital_signs
P02_ARTERIAL_MEDIAN	input	real	float64	vital_signs
P02_VENOUS_MEDIAN	input	real	float64	vital_signs
PC02_ARTERIAL_MEDIAN	input	real	float64	vital_signs
PC02_VENOUS_MEDIAN	input	real	float64	vital_signs
PCR_MEDIAN	input	real	float64	vital_signs
PH_ARTERIAL_MEDIAN	input	real	float64	vital_signs
PH_VENOUS_MEDIAN	input	real	float64	vital_signs
PLATELETS_MEDIAN	input	real	float64	vital_signs
POTASSIUM_MEDIAN	input	real	float64	vital_signs
SAT02_ARTERIAL_MEDIAN	input	real	float64	vital_signs
SAT02_VENOUS_MEDIAN	input	real	float64	vital_signs
SODIUM_MEDIAN	input	real	float64	vital_signs
TGO_MEDIAN	input	real	float64	vital_signs
TGP_MEDIAN	input	real	float64	vital_signs
TTPA_MEDIAN	input	real	float64	vital_signs
UREA_MEDIAN	input	real	float64	vital_signs
DIMER_MEDIAN	input	real	float64	vital_signs
BLOODPRESSURE_DIASTOLIC_MEAN	input	real	float64	vital_signs
BLOODPRESSURE_SISTOLIC_MEAN	input	real	float64	vital_signs
HEART_RATE_MEAN	input	real	float64	vital_signs
RESPIRATORY_RATE_MEAN	input	real	float64	vital_signs
TEMPERATURE_MEAN	input	real	float64	vital_signs
OXYGEN_SATURATION_MEAN	input	real	float64	vital_signs
BLOODPRESSURE_DIASTOLIC_MEDIAN	input	real	float64	vital_signs
BLOODPRESSURE_SISTOLIC_MEDIAN	input	real	float64	vital_signs
HEART_RATE_MEDIAN	input	real	float64	vital_signs
RESPIRATORY_RATE_MEDIAN	input	real	float64	vital_signs
TEMPERATURE_MEDIAN	input	real	float64	vital_signs
OXYGEN_SATURATION_MEDIAN	input	real	float64	vital_signs
BLOODPRESSURE_DIASTOLIC_MIN	input	real	float64	vital_signs
BLOODPRESSURE_SISTOLIC_MIN	input	real	float64	vital_signs
HEART_RATE_MIN	input	real	float64	vital_signs
RESPIRATORY_RATE_MIN	input	real	float64	vital_signs
TEMPERATURE_MIN	input	real	float64	vital_signs
OXYGEN_SATURATION_MIN	input	real	float64	vital_signs
BLOODPRESSURE_DIASTOLIC_MAX	input	real	float64	vital_signs
BLOODPRESSURE_SISTOLIC_MAX	input	real	float64	vital_signs
HEART_RATE_MAX	input	real	float64	vital_signs
RESPIRATORY_RATE_MAX	input	real	float64	vital_signs
TEMPERATURE_MAX	input	real	float64	vital_signs
OXYGEN_SATURATION_MAX	input	real	float64	vital_signs
BLOODPRESSURE_DIASTOLIC_DIFF	input	real	float64	vital_signs
BLOODPRESSURE_SISTOLIC_DIFF	input	real	float64	vital_signs
HEART_RATE_DIFF	input	real	float64	vital_signs
RESPIRATORY_RATE_DIFF	input	real	float64	vital_signs
TEMPERATURE_DIFF	input	real	float64	vital_signs
OXYGEN_SATURATION_DIFF	input	real	float64	vital_signs
BLOODPRESSURE_DIASTOLIC_DIFF_REL	input	real	float64	vital_signs
BLOODPRESSURE_SISTOLIC_DIFF_REL	input	real	float64	vital_signs
HEART_RATE_DIFF_REL	input	real	float64	vital_signs
RESPIRATORY_RATE_DIFF_REL	input	real	float64	vital_signs
TEMPERATURE_DIFF_REL	input	real	float64	vital_signs
OXYGEN_SATURATION_DIFF_REL	input	real	float64	vital_signs
GO_ICU	target	binary	int64	none

Individual Assessment¶

Student Name: Hubert Luszczyszyn (21415922)¶

OBJECTIVE¶

Info about the dataset (from author)¶

Available data:¶

Install packages¶

Load data¶

First view at the data¶

Exploratory Data Analysis (EDA) ¶

Missing Data handling¶

Create "GO_ICU" column which will indicate if patient eventually went to ICU¶

Filling up the missing data with the data of the same patient¶

Execute operation mentioned in "Important Info"¶

Removing "ICU" and "WINDOW" as, from now, they have no value for us¶

ASSUMPTION¶

Keep dtypes as they were¶

DATA DUPLICATIONS IN COLUMNS¶

Keep dtypes as they were¶

Convert "AGE_PERCENTIL" data into int64¶

METADATA¶

DATA VISUALIZATION¶

final dataset columns¶

Proportion of patients admited and not admited to ICU¶

Age distribution¶

ICU admitions proportion between patients below/above 65¶

ICU admissions proportion according to AGE_PERCENTILE¶

Create heatmap of two feature which are related to the age of a patient¶

ICU admissions based on gender¶

ICU admissions based on Hypertension¶

ICU admissions based on feature "Immunocompromised"¶

ICU admissions based on other¶

Disease gropuping¶

Create dataframe which contains vital signs data only¶

Applying standard deviation onto data to visualize difference between values and targets¶

Feature choosing¶

Strategy:¶

Drop Low variance features¶

Sort features by correlation to the target¶

Keep only 19 features which have the highest correlation with the target¶

Drop correlated columns¶

Columns which are outside top 10¶

Final view at features¶

Final shape of the dataset for training¶

evening out target imbalance (oversampling)¶

Model fitting¶

Split the data between train and test data (20% test data)¶

Testing 7 models at the same time¶

Summary of the most efficient model¶

Random Forest¶